U.S. patent application number 12/595106 was filed with the patent office on 2010-02-18 for transmission method, transmission device, receiving method, and receiving device.
This patent application is currently assigned to Naoki SUEHIRO. Invention is credited to Naoki Suehiro.
Application Number | 20100040162 12/595106 |
Document ID | / |
Family ID | 39863674 |
Filed Date | 2010-02-18 |
United States Patent
Application |
20100040162 |
Kind Code |
A1 |
Suehiro; Naoki |
February 18, 2010 |
TRANSMISSION METHOD, TRANSMISSION DEVICE, RECEIVING METHOD, AND
RECEIVING DEVICE
Abstract
A transmitting device and a receiving device wherein, on the
transmitting side, a signal creation unit creates, as its output, a
signal generated adding up the signals assuming that different data
has passed through multiple virtual channels and, on the receiving
side, oversampling is performed, the sampled data is distributed,
and signals are detected assuming that the distributed data is the
output of multiple virtual reception antennas.
Inventors: |
Suehiro; Naoki; (Ibaraki,
JP) |
Correspondence
Address: |
WESTERMAN, HATTORI, DANIELS & ADRIAN, LLP
1250 CONNECTICUT AVENUE, NW, SUITE 700
WASHINGTON
DC
20036
US
|
Assignee: |
SUEHIRO; Naoki
Tsukuba-shi, Ibaraki
JP
|
Family ID: |
39863674 |
Appl. No.: |
12/595106 |
Filed: |
March 17, 2008 |
PCT Filed: |
March 17, 2008 |
PCT NO: |
PCT/JP2008/054919 |
371 Date: |
October 8, 2009 |
Current U.S.
Class: |
375/260 ;
375/295 |
Current CPC
Class: |
H04L 27/2636 20130101;
H04J 11/00 20130101; H04L 5/0023 20130101; H04B 7/0697 20130101;
H04J 13/18 20130101; H04L 5/0007 20130101; H04L 25/0204
20130101 |
Class at
Publication: |
375/260 ;
375/295 |
International
Class: |
H04L 27/28 20060101
H04L027/28; H04L 27/00 20060101 H04L027/00 |
Foreign Application Data
Date |
Code |
Application Number |
Apr 10, 2007 |
JP |
2007-103078 |
Mar 4, 2008 |
JP |
PCT/JP2008/053866 |
Claims
1. A transmitting method that transmits a plurality of signals
generated by calculating a Kronecker product of each of sequences
that are N (N is a natural number equal to or larger than 3) row
vectors or column vectors of an Nth order DFT matrix, or of each of
N sequences of a ZCCZ sequence set, and a pilot signal and
transmission data having a length M (M is a natural number equal to
or larger than 2), said N sequences being f.sub.0, f.sub.1,
f.sub.2, . . . f.sub.N-1, said transmitting method comprising the
steps of: allocating P (P is a natural number equal to or larger
than 2) sequences out of the N sequences to pilot sequences for
transmitting pilot signals, and N-P sequences to data sequences for
transmitting transmission data; preparing Q (Q is a natural number
equal to or larger than 2 and equal to or smaller than P) pieces of
virtual channel generation data configured by two-phase signals,
four-phase signals, or complex signals; generating R (R is a
natural number equal to or larger than 1 and equal to or smaller
than Q) transmission signals each of which comprises a Kronecker
product of one of the pilot sequences and a pilot signal and a
Kronecker product of the data sequences and transmission data; and
convoluting each of the generated R transmission signals with one
separate piece of the virtual channel generation data and
transmitting the convoluted signals.
2. The transmitting method according to claim 1 wherein the pilot
sequences included in the generated transmission signals are
different among the generated transmission signals and the virtual
channel generation data used for the convolution are different
among the generated transmission signals.
3. The transmitting method according to claim 1 wherein the pilot
signal is a vector I.sub.M(1, 0, . . . 0) having a length M.
4. A transmitting device used for the transmitting method according
to claim 1, comprising: transmission signal generation units each
of which generates a transmission signal by calculating a Kronecker
product of each of sequences that are N (N is a natural number
equal to or larger than 3) row vectors or column vectors of an Nth
order DFT matrix, or of each of N sequences of a ZCCZ sequence set,
and a pilot signal and transmission data having a length M (M is a
natural number equal to 2 or larger); a data convolution unit that
convolutes R transmission signals, generated by said transmission
signal generation units, with virtual channel generation data; and
a transmitting unit that transmits the transmission signal
generated by the convolution by said data convolution unit.
5. A receiving method that receives signals transmitted by the
transmitting method according to claim 1, said receiving method
comprising: a reception step that receives the transmitted signals;
an oversampling step that performs U-fold (where U.gtoreq.R)
oversampling for the signals received by said reception step; a
channel characteristics detection step that detects channel
characteristics on a time axis for P.times.U channels; a signal
detection step that detects M.times.U reception signals; a
simultaneous equation generation step that generates M.times.R
simultaneous equations based on the channel characteristics on a
time axis for P.times.U channels and the M.times.U reception
signals; and a decoding step that solves the simultaneous equations
generated by said simultaneous equation generation step.
6. A receiving device that receives signals transmitted by the
transmitting method according to claim 1, said receiving device
comprising: a reception unit that receives the transmitted signals;
an oversampling unit that performs U-fold (where U.gtoreq.R)
oversampling for the signals received by said reception unit; a
channel characteristics detection unit that detects channel
characteristics on a time axis for P.times.U channels; a signal
detection unit that detects M.times.U reception signals; a
simultaneous equation generation unit that generates M.times.R
simultaneous equations based on the channel characteristics on a
time axis for P.times.U channels and the M.times.U reception
signals; and a decoding unit that solves the simultaneous equations
generated by said simultaneous equation generation step.
7. The transmitting method according to claim 1 wherein the R
transmission signals, each generated by the convolution with one
piece of the virtual channel generation data, are transmitted via
one or more transmission antennas.
8. The transmitting method according to claim 1 wherein the R
transmission signals, each generated by the convolution with one
piece of the virtual channel generation data, are transmitted via
different transmission antennas, one transmission signal for each
transmission antenna.
9. The transmitting method according to claim 7 wherein users are
made to correspond to transmission antennas, one user for each
transmission antenna.
10. The transmitting device according to claim 4, further
comprising an addition unit that adds up the transmission signals
generated by the convolution by said data convolution unit wherein
said transmission unit comprises one or more antennas for
transmitting the transmission signal generated by the addition by
said addition unit.
11. The transmitting device according to claim 4 wherein said
transmission unit transmits the R transmission signals via
different transmission antennas, one transmission signal for each
transmission antenna.
12. The transmitting device according to claim 10 wherein users are
made to correspond to transmission antennas, one user for each
transmission antenna.
13. The receiving method according to claim 5 wherein said
reception step receives the R transmission signals, each generated
by the convolution with one piece of the virtual channel generation
data, via one or more reception antennas.
14. The receiving method according to claim 5 wherein said
reception step receives the R transmission signals, each generated
by the convolution with one piece of the virtual channel generation
data, via different reception antennas, one transmission signal for
each reception antenna.
15. The receiving method according to claim 13 wherein users are
made to correspond to reception antennas, one user for each
reception antenna.
16. The receiving device according to claim 6 wherein said
reception unit receives the R transmission signals, each generated
by the convolution with one piece of the virtual channel generation
data, via one or more reception antennas.
17. The receiving device according to claim 6 wherein said
reception unit receives the R transmission signals, each generated
by the convolution with one piece of the virtual channel generation
data, via different reception antennas, one transmission signal for
each reception antenna.
18. The receiving device according to claim 16 wherein users are
made to correspond to reception antennas, one user for each
reception antenna.
19. The transmitting method according to claim 1 wherein when the
pilot signal is X.sub.K(x.sub.K0, x.sub.K1, x.sub.K2, . . . ,
x.sub.K(M-1)) having a length M [Mathematical Expression 34] one
sequence f.sub.0(W.sub.N.sup.0, W.sub.N.sup.0, . . . ,
W.sub.N.sup.0) out of the N sequences is allocated to the pilot
sequence for transmitting the pilot signal and N-1 sequences are
allocated to the data sequences for transmitting transmission data
and a Kronecker product of the converted pilot signal and the pilot
sequence is calculated.
20. A transmitting method that transmits a plurality of length
signals generated by calculating a Kronecker product of each of
sequences that are N (N is a natural number equal to or larger than
3) row vectors or column vectors of an Nth order DFT matrix, or of
each of N sequences of a ZCCZ sequence set, and a pilot signal and
transmission data having a length M (M is a natural number equal to
or larger than 2), said N sequences being f.sub.0, f.sub.1,
f.sub.2, . . . f.sub.N-1, said transmitting method comprising the
steps of: [Mathematical Expression 35] allocating one sequence
f.sub.0(W.sub.N.sup.0, W.sub.N.sup.0, . . . W.sub.N.sup.0) out of
the N sequences to a pilot sequence for transmitting pilot signals,
and N-1 sequences to data sequences for transmitting transmission
data; preparing Q (Q is a natural number equal to or larger than 1)
pieces of virtual channel generation data configured by two-phase
signals, four-phase signals, or complex signals; generating a first
transmission signal, which comprises signals of a Kronecker product
of the pilot sequence and the pilot signal and a Kronecker product
of the data sequences and transmission data, and a second
transmission signal which comprises signals of a Kronecker product
of the data sequences and transmission data; and when the first
transmission signal or the second transmission signal is convoluted
with one piece of the virtual channel generation data and is
transmitted, transmitting the first transmission signal once each
time the second transmission signal is transmitted (number of
virtual transmission channels-1) times.
21. A transmitting/receiving system wherein, on a transmitting
side, a signal creation unit creates a signal, which is generated
by adding up signals assuming that separate data has passed through
each of a plurality of virtual channels, as an output of said
signal creation unit and, on a receiving side, oversampling is
performed, the sampled data is distributed, and signals are
detected assuming that the distributed data is an output of a
plurality of virtual reception antennas.
22. A transmitting device wherein a signal creation unit creates a
signal generated by adding up signals assuming that separate data
has passed through each of a plurality of virtual channels and the
signal created by said signal creation unit is transmitted.
23. A receiving device wherein oversampling is performed for a
received signal, the sampled data is distributed, and signals are
detected assuming that the distributed data is an output of a
plurality of virtual reception antennas.
24. A transmitting/receiving system wherein, when separate data is
transmitted from one transmitter to each of a plurality of
receivers, a transmitter side transmits pilot signals in such a way
that the pilot signals can be separated without using channel
characteristics and transmits the separate data by inputting the
separate data into separate virtual transmission channels and
adding up the resulting data and each of said plurality of
receivers performs oversampling and distributes the sampled result
and, assuming a plurality of virtual reception antennas, generates
plural simultaneous linear equations, which can estimate
transmission data, and estimates transmission data by solving the
plural simultaneous linear equations.
25. A receiving device that receives pilot signals and data from a
plurality of transmitters, said pilot signals being transmitted so
that a pilot signal corresponding to each transmitter can be
separated without using channel characteristics, said data being
added to the pilot signals and distributes an oversampled result
generated by performing oversampling for received signals and,
assuming that the oversampled result is outputs of a plurality of
virtual reception antennas, separates the transmitters according to
channel characteristics so that plural simultaneous linear
equations can solve diversified channel characteristics between the
transmitters and the receiver, and estimates transmission data.
26. A signal transmitted from a transmitter having virtual channels
is configured by P pilot signals, P virtual channels, and a linear
combination of row (or column) vectors using the row (or column)
vectors of a matrix, that is, row vector f.sub.N,0, row vector
f.sub.N,1 . . . row vector f.sub.N,N-1 (hereinafter also called
"row vector f.sub.0, row vector f.sub.1, . . . row vector
f.sub.N-1" or "f.sub.0, f.sub.1, . . . f.sub.N-1"), and the
transmission signal is convoluted with one piece of different
virtual generation data before being transmitted. The transmission
data is transmitted, in advance, as (N-P).times.M pieces of data,
along with the pilot signal, for each virtual channel, and is
received as a reception signal wherein the pilot signal of a
virtual channel is generated by calculating the Kronecker product
of the pilot signal and one of different row (or column) vectors so
that the pilot signal can be received with no interference with
other pilots or data. On each receiving side, U (U.gtoreq.P)
virtual channels are generated, the signal received by the antenna
is branched into the U virtual channels, and the branched U signals
are processed. (1) On the transmitting side, the signals are input
to P different transmitting-side virtual channels, the outputs are
added up, and the resulting signal is transmitted over the actual
transmission channel and, on the receiving side, the received
signal is input to U (U.gtoreq.P) different virtual channels. (2)
Because there are P virtual channels on the transmitting side and U
virtual channels on the receiving side, there are PU virtual
channels via which the signal passes. Out of the PU virtual channel
characteristics, at least P.sup.2 virtual channel characteristics
are detected. (3) Using the PU channel characteristics, obtained on
the receiving side, and the data transmitted from the transmitting
side and output from the different receiving-side virtual channels,
the simultaneous linear equations are generated and the
simultaneous linear equations are solved to receive the
transmission signals with no interference. As a method for
inputting a received signal to U (U.gtoreq.P) different virtual
channels on the receiving side, a receiving method is provided that
convolutes a received signal with U different pieces of virtual
channel generation data.
Description
FIELD OF THE INVENTION
[0001] The present invention relates to a transmitting method, a
transmitting device, a receiving method, and a receiving device,
and more particularly to a transmitting method and a transmitting
device that transmit data by convoluting transmission signals with
virtual channel data, and to a receiving device.
[0002] The present invention is applicable to a wired communication
system and a wireless communication system.
[0003] A wireless communication system to which the present
invention is applied includes a wireless communication system such
as a mobile communication system and a wireless LAN communication
system.
BACKGROUND ART
[0004] Suehiro and his colleagues have devised the Suehiro's
DFT(OSDM) system that is a new information transmitting system
using the Kronecker product of the row vectors (The "column
vectors" may also be used instead of the "row vectors". The row
vectors are used in the description of this specification) of a DFT
(Discrete Fourier Transform) matrix and the data vectors (see
Non-Patent Documents 1 and 2).
[0005] It is recognized that the wireless frequency usage
efficiency of this system is about two times higher than that of
the OFDM (Orthogonal Frequency Division Multiplex) system that is
used today in various types of communication (see Non-Patent
Document 3).
[0006] Next, the following describes the OSDM (Orthogonal Signal
Division Multiplex) system that is a transmitting method for
transmitting/receiving signals having the length M.times.N
generated by calculating the Kronecker product of each of the N row
vectors (N is a natural number equal to or larger than 3) of an Nth
order DFT matrix and data having the length M (M is a natural
number equal to or larger than 2).
(DFT Matrix and Transmission Signal)
[0007] First, the following describes an Nth order DFT (Discrete
Fourier Transform) matrix.
[0008] Let the Nth order DFT matrix F.sub.N be defined as
follows.
F.sub.N=[f.sub.N(i,j)] (1)
[0009] where the Nth order inverse DFT matrix F.sub.N.sup.-1 is the
complex conjugate of the DFT matrix F.sub.N.
[0010] In the above expression, i is a row number
(0.ltoreq.i.ltoreq.N-11) and j is a column number
(0.ltoreq.j.ltoreq.N-1).
f.sub.N(i,j)=exp(2.pi. -1ij/N)/ N (2)
[0011] The variable W.sub.N corresponding to the point, generated
by dividing a unit circle into N, is defined as follows as shown in
FIG. 1.
W.sub.N.ident.exp(2.pi. -1)/N (3)
[0012] When this W.sub.N is used, the DFT matrix F.sub.N is as
shown in FIG. 2.
[0013] W.sub.N is a rotor and the following relation is
satisfied.
W.sub.N.sup.N=e.sup.j2.pi.=1 (4)
W.sub.N.sup.N-k=W.sub.N.sup.2N-k= . . . =W.sub.N.sup.-k (5)
As shown in FIG. 2, the Nth order DFT matrix F.sub.N has N row
vectors, that is, vector f.sub.N,0, vector f.sub.N,1 . . . vector
f.sub.N,N-1. The cyclic crosscorrelations among those row vectors
are zero in all shifts.
[0014] Next, the following describes data transmission using those
row vectors. As shown in FIG. 3, the signals S.sub.0, S.sub.1 . . .
S.sub.N-1 are generated from N pieces of transmission data each
having the length M (data X.sub.0(x.sub.00, x.sub.01, . . . ,
x.sub.0(M-1)), data X.sub.1(x.sub.10, x.sub.11, . . . ,
x.sub.1(M-1)), . . . , data X.sub.(N-1)(x.sub.(N-1)0, x.sub.(N-1)1,
. . . , x.sub.(N-1)(M-1))), received from transmitting unit #0,
transmitting unit #1 . . . transmitting unit #(N-1) using row
vector f.sub.N,0, row vector f.sub.N,1, . . . , row vector
f.sub.N,N-1, and the generated signals are transmitted.
S 0 = f N , 0 DataX 0 ##EQU00001## S 1 = f N , 1 DataX 1
##EQU00001.2## ##EQU00001.3## S N - 1 = f N , N - 1 DataX ( N - 1 )
##EQU00001.4##
[0015] where is the Kronecker product.
Transmitting the generated signals S.sub.0, S.sub.1 . . . S.sub.N-1
allows data to be transmitted from multiple transmitting units
without correlation. Note that the length of transmitted signals is
N.times.M.
[0016] That is, because the cyclic crosscorrelation between any two
signals of the signals S.sub.0, S.sub.1 . . . S.sub.N-1 is zero in
all shifts, the well-designed matched filters allows data sequences
to be separated at reception time even when the signals are added
up.
(Matched Filter)
[0017] The vector I.sub.M(1, 0, . . . , 0) having the length M is
defined.
[0018] Here, the matched filters for matching to the signals of the
Kronecker product of the vectors f.sub.k (0.ltoreq.k.ltoreq.N-1)
and I.sub.M are provided.
[Mathematical Expression 2]
f.sub.kI.sub.M=(W.sub.N.sup.0,0, . . . , 0,W.sub.N.sup.k,0, . . . ,
0,W.sub.N.sup.(N-1)k,0, . . . , 0)/ N (7)
[0019] When the signals S.sub.k (0.ltoreq.k.ltoreq.N-1) are input
to the matched filters, M units of data in the center of the output
becomes data X.sub.K.
[0020] In addition, when the signals S.sub.g (where g.noteq.k,
0.ltoreq.k.ltoreq.N-1, 0.ltoreq.g.ltoreq.N-1) are input to the
matched filters
[Mathematical Expression 3]
f.sub.kI.sub.M=(W.sub.N.sup.0,0, . . . , 0,W.sub.N.sup.k,0, . . . ,
0,W.sub.N.sup.(N-1)k,0, . . . , 0)/ N (8)
[0021] The M units of data in the center of the output signal are
always 0. This means that, even when the signals, from signal
S.sub.0 to signal S.sub.N-1, are added up, only X.sub.K is produced
when they are input to the matched filters of f.sub.KI.sub.M.
(Pseudo-Periodic Signal)
[0022] Let S.sub.sum be the signal produced by adding up signal
S.sub.0 to signal S.sub.N-1. Because the signal S.sub.sum is a
limited-length sequence having the length MN, the periodicity
obtained by the DFT matrix is lost when the signal is input to the
multipath channels. In such a case, the data X.sub.k
(0.ltoreq.k.ltoreq.N-1) cannot be obtained from the matched filter
output.
[0023] Multipath channels do not affect the periodicity of the
signal if the signal is a periodic signal having an unlimited
length. However, transmitting a sequence having an unlimited length
is not practical. To solve this problem, a pseudo-periodic signal,
generated by selecting a signal having a necessary length from the
periodic sequence of an unlimited length, is used.
[0024] First, let L.sub.2 be a value larger than the assumed
multipath delay time.
[0025] When there is no direct-path signal or when the power level
of the direct-path signal is extremely low, the delay time for the
maximum amplitude signal becomes sometimes negative. Let L.sub.1 be
a value considering that time.
[0026] Using those values L.sub.1 and L.sub.2, the pseudo-periodic
signal, such as the one shown in FIG. 4, is generated and
transmitted.
[0027] The part corresponding to L.sub.2 is called a cyclic prefix,
and the part corresponding to L.sub.1 is called a cyclic postfix.
At reception time, both prefixes must be removed before the signal
enters the matched filter.
(Pilot Signal)
[0028] The data sequence X.sub.0 is defined as follows where the
length is M.
X.sub.0=(1,0,0,0, . . . , 0) (9)
[Mathematical Expression 4]
[0029] When "f.sub.kI.sub.M" is calculated using this data sequence
and the data sequence is input directly into f.sub.kI.sub.M, the
central part of the output becomes as follows.
X.sub.0=(1,0,0,0, . . . , 0) (10)
[0030] Next, S.sub.0 is converted to a pseudo-periodic signal,
which is sent via multipath channels. When the cyclic (pre/post)
prefixes are removed and the signal is input to the matched filters
of f.sub.kI.sub.M, the M units of data in the central part of the
output are as follows.
X0=(p0,p1,p2,p3, . . . , p(L2-1),0,0, . . . , 0) (11)
[0031] where (p.sub.0, p.sub.1, p.sub.2, p.sub.3, . . . , p.sub.k,
. . . , p.sub.(L2-1)) are complex coefficients that are multiplied
by the paths which arrived with a delay of time k. They correspond
to the transmission characteristics including the transmission
characteristics of the transmitting device, the transmission
characteristics of the propagation space, and the transmission
characteristics of the receiving device and represent the channel
characteristics on the time axis.
[0032] This p.sub.k is usually represented as shown below using the
amplitude coefficient r.sub.k and the phase rotation
.theta..sub.k.
p.sub.k=r.sub.ke.sup.j.theta.k (12)
[0033] As the pilot signal, the signal of the ZACZ (Zero Auto
Correlation Zone Sequence) sequence, the signal of the ZCCZ (Zero
Crosscorrelation Zone Sequence) sequence, and the signal of the PN
sequence may be used.
[Mathematical Expression 5]
[0034] In this case, the output of the matched filters of
f.sub.kI.sub.M described above must be input to the matched filters
that match to those pilot signals.
[0035] Also when ZACZ and so on are used as the pilot signal, the
channel characteristics on the time axis, including the multipath
characteristics, may be detected.
(Simultaneous Equation)
[0036] As described above, the channel characteristics on the time
axis, including the multipath characteristics, can be obtained by
inserting the pilot signal.
[0037] The M units of data (dk0-dk(M-1)) in the center of each
matched filter output of the data signal parts Xk (1<k<N-1),
other than the pilot, have the relation between the data and the
multipath characteristics which is shown by the following
expression.
( p 0 , p 1 , , p L 2 - 2 , p L 2 - 1 , 0 , , 0 , 0 , 0 ) x k 0 + (
0 , p 0 , p 1 , , p L 2 - 2 , p L 2 - 1 , 0 , , 0 , 0 ) x k 1 + ( 0
, 0 , p 0 , p 1 , , p L 2 - 2 , p L 2 - 1 , 0 , , 0 ) x k 2 + ( 0 ,
0 , 0 , , 0 , 0 , 0 , 0 , p 0 , p 1 ) x k ( M - 2 ) + ( 0 , 0 , 0 ,
0 , , 0 , 0 , 0 , 0 , p 0 ) x k ( M - 1 ) = ( d k 0 , d k 1 , d k 2
, , d k ( M - 2 ) , d k ( M - 1 ) ) ( 13 ) ##EQU00002##
This is expressed by the matrix shown in Expression (14) given
below.
[ Mathematical expression 6 ] [ d k 0 d k 1 d k 2 d k ( L 2 - 1 ) d
kL 2 d k ( M - 1 ) ] = [ P 0 0 0 0 0 P 1 P 1 P 0 0 0 0 P 2 P 2 P 1
P 0 0 0 P 3 P L 2 - 1 P L 2 - 2 P L 2 - 3 P 0 0 0 0 P L 2 - 1 P L 2
- 2 P 1 P 0 0 0 0 0 0 0 P 0 ] [ I k 0 I k 1 I k 2 I k ( L 2 - 1 ) I
kL 2 I k ( M - 1 ) ] ( 14 ) ##EQU00003##
where
[ Mathematical expression 7 ] If P = [ P 0 0 0 0 0 P 1 P 1 P 0 0 0
0 P 2 P 2 P 1 P 0 0 0 P 3 P L 2 - 1 P L 2 - 2 P L 2 - 3 P 0 0 0 0 P
L 2 - 1 P L 2 - 2 P 1 P 0 0 0 0 0 0 0 P 0 ] ( 15 ) D k = [ d k 0 d
k 1 d k 2 d k ( L 2 - 1 ) d kL 2 d k ( M - 1 ) ] ( 16 )
##EQU00004##
then,
D.sub.k=P.sup.tX.sub.k (17)
[0038] Solving Expression (17) for X.sub.k allows the receiving
side to obtain the transmission data for which the channel
characteristics on the time axis are compensated, wherein the
channel characteristics include the transmission characteristics of
the transmitting device side, the transmission characteristics of
the propagation space, and the transmission characteristics of the
receiving device side.
[0039] To solve this simultaneous equation simply, the both sides
of Expression (17) are multiplied by the inverse matrix of P from
the left.
P.sup.-1D.sub.k=P.sup.-1P.sup.tX.sub.k
=.sup.tX.sub.k (18)
Non-Patent Document 1: N. Suehiro, C. Han, T. Imoto, and N.
Kuroyanagi, "An information transmission method using Kronecker
product" Proceedings of the IASTED International Conference
Communication Systems and Networks, pp. 206-209, September 2002.
Non-Patent Document 2: N. Suehiro, C. Han, and T. Imoto, "Very
Efficient wireless usage based on pseudo-coherent addition of
multipath signals using Kronecker product with rows of DFT matrix",
Proceedings of International Symposium on Information Theory, pp.
385, June 2003. Non-Patent Document 3: Naoki Suehiro, Rongzhen Jin,
Chenggao Han, Takeshi Hashimoto, "Performance of Very Efficient
Wireless Frequency Usage System Using Kronecker Product with Rows
of DFT Matrix", Proceedings of 2006 IEEE Information Theory
Workshop (ITW'06) pp. 526-529, October 2006.
DISCLOSURE OF THE INVENTION
Problems to be Solved by the Invention
[0040] However, though the wireless frequency usage efficiency of
the conventional OSDM system is about two times higher than that of
the OFDM system, the problem to be solved is to further increase
the wireless frequency usage efficiency.
[0041] In view of the foregoing, it is an object of the present
invention to provide a transmitting method, a transmitting device,
a receiving method, and a receiving device that have higher
wireless frequency usage efficiency.
Means to Solve the Problems
[0042] To achieve the above objects, a transmitting method of the
present invention is a transmitting method that transmits a
plurality of signals generated by calculating a Kronecker product
of each of sequences that are N (N is a natural number equal to or
larger than 3) row vectors or column vectors of an Nth order DFT
matrix, or of each of N sequences of a ZCCZ sequence set, and a
pilot signal and transmission data having a length M (M is a
natural number equal to or larger than 2), the N sequences being
f.sub.0, f.sub.1, f.sub.2, . . . f.sub.N-1, the transmitting method
comprising the steps of allocating P (P is a natural number equal
to or larger than 2) sequences out of the N sequences to pilot
sequences for transmitting pilot signals, and N-P sequences to data
sequences for transmitting transmission data; preparing Q (Q is a
natural number equal to or larger than 2 and equal to or smaller
than P) pieces of virtual channel generation data configured by
two-phase signals, four-phase signals, or complex signals;
generating R (R is a natural number equal to or larger than 1 and
equal to or smaller than Q) transmission signals each of which
comprises a Kronecker product of one of the pilot sequences and a
pilot signal and a Kronecker product of the data sequences and
transmission data; and convoluting each of the generated R
transmission signals with one separate piece of the virtual channel
generation data and transmitting the convoluted signals.
[0043] In the present invention, the linear combination of the row
vectors of a DFT matrix and the linear combination of the column
vector of a DFT matrix also have the equivalent function of the row
vectors of a DFT matrix and the column vectors of a DFT matrix.
Therefore, in the present invention, the row vectors of a DFT
matrix and the column vectors of a DFT matrix each include the
linear combination of the row vectors of a DFT matrix and the
linear combination of the column vectors of a DFT matrix.
[0044] To achieve the above objects, a transmitting device of the
present invention is a transmitting device comprising transmission
signal generation units each of which generates a transmission
signal by calculating a Kronecker product of each of sequences that
are N (N is a natural number equal to or larger than 3) row vectors
or column vectors of an Nth order DFT matrix, or of each of N
sequences of a ZCCZ sequence set, and a pilot signal having a
length M (M is a natural number equal to 2 or larger) and
transmission data having a length M; a data convolution unit that
convolutes R transmission signals, generated by the transmission
signal generation units, with virtual channel generation data; and
a transmitting unit that transmits the transmission signal
generated by the convolution by the data convolution unit.
[0045] To achieve the above objects, a receiving method of the
present invention is a receiving method that receives signals
transmitted by the transmitting method according to one of claims
1-3, the receiving method comprising a reception step that receives
the transmitted signals; an oversampling step that performs U-fold
(where U.gtoreq.R) oversampling for the signals received by the
reception step; a channel characteristics detection step that
detects channel characteristics on a time axis for P.times.U
channels; a signal detection step that detects M.times.U reception
signals; a simultaneous equation generation step that generates
M.times.R simultaneous equations based on the channel
characteristics on a time axis for P.times.U channels and the
M.times.U reception signals; and a decoding step that solves the
simultaneous equations generated by the simultaneous equation
generation step.
[0046] To achieve the above objects, a receiving device of the
present invention is a receiving device that receives signals
transmitted by the transmitting method according to one of claims
1-3, the receiving device comprising a reception unit that receives
the transmitted signals; an oversampling unit that performs U-fold
(where U.gtoreq.R) oversampling for the signals received by the
reception unit; a channel characteristics detection unit that
detects channel characteristics on a time axis for P.times.U
channels; a signal detection unit that detects M.times.U reception
signals; a simultaneous equation generation unit that generates
M.times.R simultaneous equations based on the channel
characteristics on a time axis for P.times.U channels and the
M.times.U reception signals; and a decoding unit that solves the
simultaneous equations generated by the simultaneous equation
generation step.
EFFECTS OF THE INVENTION
[0047] The present invention provides a transmitting method, a
transmitting device, a receiving method, and a receiving device
that have higher wireless frequency usage efficiency.
BRIEF DESCRIPTION OF THE DRAWINGS
[0048] FIG. 1 is a diagram showing W.sub.N.
[0049] FIG. 2 is a diagram showing an Nth order DFT matrix.
[0050] FIG. 3 is a diagram showing the transmission of a signal
using the row vectors of the DFT matrix F.sub.N.
[0051] FIG. 4 is a diagram showing a pseudo-periodic signal.
[0052] FIG. 5 is a diagram showing the configuration of the signal
of transmission signal 0.
[0053] FIG. 6 is a diagram showing the configuration of the signal
of transmission signal 1.
[0054] FIG. 7 is a diagram showing the configuration of the signal
of transmission signal P-31 1.
[0055] FIG. 8 is a diagram showing an example of virtual channel
generation data.
[0056] FIG. 9 is a diagram showing a transmitting device (1).
[0057] FIG. 10 is a diagram showing a virtual channel generation
data convolution and addition unit.
[0058] FIG. 11 is a diagram showing convolution using virtual
channel generation data.
[0059] FIG. 12 is a diagram showing a receiving device (1).
[0060] FIG. 13 is a diagram schematically showing oversampling.
[0061] FIG. 14 is a diagram (1) showing virtual channels based on
oversampling.
[0062] FIG. 15 is a diagram (2) showing virtual channels based on
oversampling.
[0063] FIG. 16 is a diagram showing a simultaneous equation
generation unit.
[0064] FIG. 17 is a diagram showing a transmitting device (2).
[0065] FIG. 18 is a diagram showing a transmitting device (3).
[0066] FIG. 19 is a diagram showing a receiving device (2).
[0067] FIG. 20 is a diagram showing a receiving device (3).
[0068] FIG. 21 is a diagram showing the effect of multipath
characteristics.
[0069] FIG. 22 is a diagram showing the transmission of pilot
signals.
[0070] FIG. 23 is a diagram showing a transmitting system (1).
[0071] FIG. 24 is a diagram showing a receiving system (1).
[0072] FIG. 25 is a diagram showing the simulation result (1).
[0073] FIG. 26 is a diagram showing the simulation result (2).
[0074] FIG. 27 is a diagram showing the simulation result (3).
[0075] FIG. 28 is a diagram showing the simulation result (4).
[0076] FIG. 29 is a diagram showing a transmitting system (2).
[0077] FIG. 30 is a diagram showing the signal configuration
(1).
[0078] FIG. 31 is a diagram showing the signal configuration
(2).
[0079] FIG. 32 is a diagram showing a receiving system (2).
[0080] FIG. 33 is a diagram showing the simulation result (5).
[0081] FIG. 34 is a diagram showing the simulation result (6).
[0082] FIG. 35 is a diagram showing the simulation result (7).
[0083] FIG. 36 is a diagram showing the simulation result (8).
[0084] FIG. 37 is a diagram showing the simulation result (9).
[0085] FIG. 38 is a diagram showing the simulation result (10).
[0086] FIG. 39 is a diagram showing virtual transmission
antennas.
[0087] FIG. 40 is a diagram showing virtual reception antennas.
[0088] FIG. 41 is a diagram showing a transmitting system (3).
[0089] FIG. 42 is a diagram showing a receiving system (3).
[0090] FIG. 43 is a diagram showing the configuration of a
transmission signal.
[0091] FIG. 44 is a diagram showing a receiving device (4)
EXPLANATIONS OF SYMBOLS
[0092] 11 Virtual channel 0 (virtual transmission channel 0)
transmission signal creation unit [0093] 12 Virtual channel 1
(virtual transmission channel 1) transmission signal creation unit
[0094] 13 Virtual channel 2 (virtual transmission channel 2)
transmission signal creation unit [0095] 15 Virtual channel
generation data convolution and addition unit [0096] 17,172
Transmitting unit [0097] 18,181,182,183,184,185 Transmission
antenna [0098] 21,211,212,213,214,215 Reception antenna [0099] 22
Receiving unit [0100] 151,381 Virtual channel generation data
storage unit [0101] 152,153,154 Convolution unit [0102] 155,1551
Addition unit [0103] 221,222,223,224,225 Receiving unit [0104] 25
Channel characteristics detection unit [0105] 26 Simultaneous
equation generation unit [0106] 27 Decoding unit [0107] 28
Oversampling unit [0108] 29 Signal separation unit [0109] 38
Receiving side virtual channel convolution unit
MOST PREFERRED MODE FOR CARRYING OUT THE INVENTION
Signal Configuration
[0110] Transmission data has the signal configuration shown in FIG.
5 to FIG. 7.
[0111] In the signal configuration in FIG. 5 to FIG. 7, the N row
vectors (N sequences) of an Nth order DFT matrix are f.sub.N,0,
f.sub.N,1, f.sub.N,2, . . . f.sub.N,N-1. Out of the N row vectors,
P row vectors f.sub.N,0-f.sub.N,P-1, are used as pilot row vectors
for transmitting pilot signals, and N-P row vectors
f.sub.N,P-f.sub.N,N-1 are used as data transmission row vectors for
transmission data.
[0112] The signal configuration in FIG. 5 shows the case in which
one row vector f.sub.N,0 is used as the pilot row vector and N-P
row vectors f.sub.N,P-f.sub.N,N-1 are used as data row vectors.
[0113] The row vectors f.sub.N,0, f.sub.N,1, f.sub.N,2, . . .
f.sub.N,N-1 may also be N column vectors of an Nth order DFT
matrix.
[0114] The vectors f.sub.N,0, f.sub.N,1, f.sub.N,2, . . .
f.sub.N,N-1 may be N sequences configuring a ZCCZ sequence set
instead of the N row vectors of an Nth order DFT matrix.
[0115] For example, the row vectors of the ZCCZ matrix defined
below may be used as the ZCCZ sequence set.
[0116] The ZCCZ matrix mentioned here is a matrix of N rows and K
columns, and a zero crosscorrelation zone exists in the k-period
cyclic crosscorrelation function between any two row vectors.
[0117] Note that, depending upon the combination of two row
vectors, a zero crosscorrelation zone may exist in the k'-period
cyclic crosscorrelation function (k' is not k (k'.noteq.k)).
[0118] The pilot signal X.sub.o(x.sub.00, x.sub.01, . . . ,
x.sub.0(M-1)) may be X.sub.0=(1, 0, 0, 0, . . . 0), or the ZCZ
sequence signal having the length M or the ZCCZ sequence signal
having the length M may be used.
[0119] The Kronecker product of each of N-P pieces of transmission
data X.sub.0,P(x.sub.0,P,0, x.sub.0,P,1, . . . , x.sub.0,P,(M-1)) .
. . X.sub.0,N-1(x.sub.0,(N-1),0, x.sub.0,(N-1),1, . . . ,
x.sub.0,(N-1),(M-1)) and N-P row vectors f.sub.N,P to f.sub.N,N-1
is calculated.
[0120] Therefore, "transmission signal 0" shown in FIG. 5 is the
following signal.
[Mathematical Expression 8]
Vector f.sub.N,0Pilot signal X.sub.0+Vector f.sub.N,PTransmission
data X.sub.0,P . . . +Vector f.sub.N,N-2Transmission data
X.sub.0,N-2+Vector f.sub.N,N-1Transmission data X.sub.0,N-1
(19)
[0121] In the signal configuration in FIG. 6, one row vector
f.sub.N,1 is used as the pilot row vector and N-P row vectors
f.sub.N,P-f.sub.N,N-1 are used as data row vectors.
[0122] "Transmission signal 1" shown in FIG. 6 is the following
signal.
[Mathematical Expression 9]
Vector f.sub.N,1Pilot signal X.sub.1+Vector f.sub.N,PTransmission
data X.sub.1,P . . . +Vector f.sub.N,N-2Transmission data
X.sub.1,N-2+Vector f.sub.N,N-1Transmission data X.sub.1,N-1
(20)
[0123] Similarly, in the signal configuration in FIG. 7, one row
vector f.sub.N,P-1 is used as the pilot row vector and N-P row
vectors f.sub.N,P-f.sub.N,N-1 are used as data row vectors.
[0124] "Transmission signal P-1" shown in FIG. 7 is the following
signal.
[Mathematical Expression 10]
Vector f.sub.N,P-1Pilot signal X.sub.P-1+Vector
f.sub.N,PTransmission data X.sub.N-1,P . . . +Vector
f.sub.N,N-2Transmission data X.sub.N-1,N-2+Vector
f.sub.N,N-1Transmission data X.sub.N-1,N-1 (21)
Virtual Channel Data
[0125] The data in FIG. 8 may be used as an example of virtual
channel generation data.
[0126] A low correlation sequence or random numbers may also be
used as the virtual channel generation data.
[0127] A virtual channel is generated by convoluting data with one
of different virtual channel generation data described above and
transmitting the convoluted data. This virtual channel is also
called a virtual transmission channel or a virtual transmission
antenna because the virtual channel is generated on the
transmitting side.
Transmitting Device
[0128] Referring to FIG. 9, the following describes a transmitting
device where there is one actual antenna and P is "3" (there are
three pilot signals and three virtual channels (virtual
transmission channel, virtual transmission antenna)) in FIG. 5-FIG.
7.
[0129] In the transmitting device in FIG. 9, the N row vectors (N
sequences) of an Nth order DFT matrix are f.sub.N,0, f.sub.N,1,
f.sub.N,2, . . . f.sub.N,N-1 and, out of those N row vectors, three
row vectors f.sub.N,0-f.sub.N,2 are used as pilot row vectors and
N-P(N3) row vectors f.sub.N,3-f.sub.N,N-1 are used as data row
vectors for transmitting transmission data.
[0130] The transmitting device in FIG. 9 comprises a virtual
channel 0 (virtual transmission channel 0, virtual transmission
antenna 0) transmission signal creation unit 11, a virtual channel
1 (virtual transmission channel 1, virtual transmission antenna 1)
transmission signal creation unit 12, a virtual channel 2 (virtual
transmission channel 2, virtual transmission antenna 2)
transmission signal creation unit 13, a virtual channel generation
data convolution and addition unit 15, a transmitting unit 17, and
an antenna 18. The virtual channel generation data convolution and
addition unit 15 comprises a virtual channel generation data
storage unit 151 in which virtual channel generation data is
stored.
[0131] The virtual channel 0 transmission signal creation unit 11
calculates the Kronecker product of the pilot row vector f.sub.N,0
and the pilot signal X.sub.0(x.sub.00, x.sub.01, . . . ,
x.sub.0(M-1)) and the Kronecker product of each of N-3 data vectors
f.sub.N,3-f.sub.N,N-1 and the N-3 pieces of transmission data
X.sub.0,3(x.sub.0,3,0, x.sub.0,3,1, . . . , x.sub.0,3,(M-1)) . . .
X.sub.0,N-1(x.sub.0,(N-1),0, x.sub.0,(N-1),1, . . . ,
x.sub.0,(N-1),(M-1)) to create virtual channel 0 transmission
signal KS0.
[0132] Finally, the virtual channel 0 transmission signal creation
unit 11 creates the following signal.
[Mathematical Expression 11]
Transmission signal KS0: Vector f.sub.N,0Pilot signal
X.sub.0+Vector f.sub.N,3Transmission data X.sub.0,3 . . . +Vector
f.sub.N,N-2Transmission data X.sub.0,N-2+Vector
f.sub.N,N-1Transmission data X.sub.0,N-1 (22)
[0133] Note that the transmission signal KS0 is the sum of (N-2)
signals having the length NM.
[0134] Similarly, the virtual channel 1 transmission signal
creation unit 12 calculates the Kronecker product of the pilot row
vector f.sub.N,1 and the pilot signal X.sub.1(x.sub.10, x.sub.11, .
. . , x.sub.1(M-1)) and the Kronecker product of each of N-3 N-3
data vectors f.sub.N,3-f.sub.N,N-1 and the N-3 pieces of
transmission data X.sub.1,3(x.sub.1,3,0, x.sub.1,3,1, . . . ,
x.sub.1,3,(M-1)) . . . X.sub.1,N-1(x.sub.1,(N-1),0,
x.sub.1,(N-1),1, . . . , x.sub.1,(N-1),(M-1)) to create the
transmission signal KS1 for virtual channel 1.
[0135] Finally, the virtual channel 1 transmission signal creation
unit 12 creates the following signal.
[Mathematical Expression 12]
Transmission signal KS1: Vector f.sub.N,1Pilot signal
X.sub.1+Vector f.sub.N,3Transmission data X.sub.1,3 . . . +Vector
f.sub.N,N-2Transmission data X.sub.1,N-2+Vector
f.sub.N,N-1Transmission data X.sub.1,N-1 (23)
[0136] Similarly, the virtual channel 2 transmission signal
creation unit 13 calculates the Kronecker product of the pilot row
vector f.sub.N,2 and the pilot signal X.sub.2(x.sub.20, x.sub.21, .
. . , x.sub.2(M-1)) and the Kronecker product of each of N-3 data
vectors f.sub.N,3-f.sub.N,N-1 and the N-3 pieces of transmission
data X.sub.2,3(x.sub.2,3,0, x.sub.2,3,1, . . . , x.sub.2,3,(M-1)) .
. . X.sub.2,N-1(x.sub.2,(N-1),0, x.sub.2,(N-1),1, . . . ,
x.sub.2,(N-1),(M-1)) to create the transmission signal KS2 for
virtual channel 2.
[0137] Finally, the virtual channel 2 transmission signal creation
unit 13 creates the following signal.
[Mathematical Expression 13]
Transmission signal KS2: Vector f.sub.N,2Pilot signal
X.sub.2+Vector f.sub.N,3Transmission data X.sub.2,3 . . . +Vector
f.sub.N,N-2Transmission data X.sub.2,N-2+Vector
f.sub.N,N-1Transmission data X.sub.2,N-1 (24)
[0138] The virtual channel generation data convolution and addition
unit 15 performs the convolution between the virtual channel 0
transmission signal KS0, virtual channel 1 transmission signal KS1,
or virtual channel 2 transmission signal KS2, respectively, and
different virtual channel data, adds the them up, and supplies the
result to the transmitting unit.
[0139] The transmitting unit 17 and the antenna 18 transmit the
signal, generated by the virtual channel generation data
convolution and addition unit 15, at high frequencies via the
antenna 18.
[0140] FIG. 10 shows an example of the virtual channel generation
data convolution and addition unit 15. The virtual channel
generation data convolution and addition unit 15 in FIG. 10
comprises the virtual channel generation data storage unit 151,
convolution units 152-154, and an addition unit 155.
[0141] The convolution unit 152 performs the convolution between
the virtual channel 0 transmission signal KS0 and virtual channel
generation data D2, the convolution unit 153 performs the
convolution between the virtual channel 1 transmission signal KS1
and virtual channel generation data D1, and the convolution unit
154 performs the convolution between the virtual channel 2
transmission signal KS2 and virtual channel generation data D0.
[0142] The addition unit 155 adds up the signals from the
convolution units 152-154 and outputs the result to the
transmitting unit 17.
[0143] Meaning of convolution units: The transmission signal KS0,
transmission signal KS1, and transmission signal KS2, added up by
the addition unit 155, are transmitted via the transmission unit
and the antenna. Because the transmission signal KS0, transmission
signal KS1, and transmission signal KS2 are transmitted through the
same space and are received by a receiver, the channel
characteristics in the space are the same. However, because the
transmission signal KS0, transmission signal KS1, and transmission
signal KS2 are convoluted with different virtual channel data, the
transmission signal KS0, transmission signal KS1, and transmission
signal KS2 are equivalent to the signals received via different
lines when viewed from the receiver.
[0144] This means that the receiving side detects the channel
characteristics of each line, generates a simultaneous equation,
and solves this simultaneous equation to correctly receive the
transmission signal KS0, transmission signal KS1, and transmission
signal KS2.
[0145] The following describes the convolution unit 152 with
reference to FIG. 11.
[0146] In the description below, let the virtual channel data D2
for the virtual channel 0 transmission signal KS0 be (1j1-j) and
let the virtual channel 0 transmission signal KS0 be (KS0.sub.0,
KS0.sub.1, KS0.sub.2, KS0.sub.3, . . . KS0.sub.NM-1).
[0147] As shown in FIG. 11, the signal generated by adding up
(KS0.sub.0, KS0.sub.1, KS0.sub.2, KS0.sub.3, . . . KS0.sub.NM-1),
j(KS0.sub.0, KS0.sub.1, KS0.sub.2, KS0.sub.3, . . . KS0.sub.NM-1)
after one time slot, (KS0.sub.0, KS0.sub.1, KS0.sub.2, KS0.sub.3, .
. . KS0.sub.NM-1) after another one time slot, and -j(KS0.sub.0,
KS0.sub.1, KS0.sub.2, KS0.sub.3, . . . KS0.sub.NM-1) after still
another one time slot is output from the convolution unit 152.
[0148] Similarly, when the virtual channel data D2 for the virtual
channel 1 transmission signal KS1 is (j11j), the signal generated
by adding up j(KS1.sub.0, KS1.sub.1, KS1.sub.2, KS1.sub.3, . . .
KS1.sub.NM-1), (KS1.sub.0, KS1.sub.1, KS1.sub.2, KS1.sub.3, . . .
KS1.sub.NM-1) after one timeslot, (KS1.sub.0, KS1.sub.1, KS1.sub.2,
KS1.sub.3, . . . KS1.sub.NM-1) after another one time slot, and
j(KS1.sub.0, KS1.sub.1, KS1.sub.2, KS1.sub.3, . . . KS1.sub.NM-1)
after still another one time slot is output from the convolution
unit 153.
[0149] Similarly, when the virtual channel generation data D0 for
the virtual channel 2 transmission signal KS2 is (1jj1), the signal
generated by adding up (KS2.sub.0, KS2.sub.1, KS2.sub.2, KS2.sub.3,
. . . KS2.sub.NM-1), j(KS2.sub.0, KS2.sub.1, KS2.sub.2, KS2.sub.3,
. . . KS2.sub.NM-1) after one time slot, j(KS2.sub.0, KS2.sub.1,
KS2.sub.2, KS2.sub.3, . . . KS2.sub.NM-1) after another one time
slot, and (KS2.sub.0, KS2.sub.1, KS2.sub.2, KS2.sub.3, . . .
KS2.sub.NM-1) after still another one time slot is output from the
convolution unit 154.
Receiving Device
[0150] The following describes a receiving device that receives the
signals transmitted from the transmitting device in FIG. 9 whose
transmission data amount is increased (N-P) times by the virtual
channels on the transmitting side.
[0151] FIG. 43 shows a generalized signal transmitted from a
transmitter having the virtual channels shown in FIG. 9.
[0152] The signal configuration shown in FIG. 43 is that there are
P pilot signals, there are P virtual channels, and the row vectors
of a DFT matrix are the row vector f.sub.N,0, row vector f.sub.N,1
. . . row vector f.sub.N,N-1 such as the those defined in FIG. 4
(hereinafter called "row vector f.sub.0, row vector f.sub.1 . . .
row vector f.sub.N-1" or "f.sub.0, f.sub.1, . . . f.sub.N-1").
[0153] For each virtual channel, the pilot signal and (N-P).times.M
pieces of data are transmitted as the transmission data.
[0154] From virtual channel #0, the Kronecker product of the row
vector f.sub.0 and the pilot signal is calculated and the pilot
signal #0 is transmitted. At the same time, from virtual channel
#0, the Kronecker product of the N-P pieces of transmission data
(X.sub.00, X.sub.01, . . . , X.sub.0(N-P)) and each of the row
vector f.sub.0, row vector f.sub.1 . . . row vector f.sub.N-1 is
calculated and the result is transmitted as transmission data
#0.
[0155] Because the length of each of the N-P pieces of transmission
data is M, M.times.(N-P) data is transmitted from virtual channel
#0.
[0156] Similarly, from virtual channel #1, the Kronecker product of
the row vector f.sub.1 and the pilot signal is calculated and the
pilot signal #1 is transmitted. At the same time, from virtual
channel #1, the Kronecker product of the N-P pieces of transmission
data (X.sub.10, X.sub.11, . . . , X.sub.1(N-P)) and each of the row
vector f.sub.0, row vector f.sub.1 . . . row vector f.sub.N-1 is
calculated and the result is transmitted as transmission data
signal #1.
[0157] Similarly, from virtual channel #P- 1, the Kronecker product
of the row vector f.sub.P-1 and the pilot signal is calculated and
the pilot signal #P- 1 is transmitted. At the same time, from
virtual channel #P-1, the Kronecker product of the N-P pieces of
transmission data (X.sub.P-10, X.sub.P-11, . . . , X.sub.P-1(N-P))
and each of the row vector f.sub.0, row vector f.sub.1 . . . row
vector f.sub.N-1 is calculated and the result is transmitted as
transmission data signal #N-1.
[0158] Because different row vectors are used for the calculation
of the Kronecker products, the pilot signal of each virtual channel
can be received with no interference from other pilots and
data.
[0159] However, the transmission data of the virtual channels share
the N-P row vectors f.sub.0-f.sub.N-1 in calculating the Kronecker
product.
[0160] As a result, if the receiving side takes no action,
interference occurs in the MP pieces of data transmitted using the
same row vector.
[0161] To solve this problem, the present invention provides a
method that generates U (U.gtoreq.P) virtual channels on the
receiving side, branches the signals received via the antenna into
the U virtual channels, processes the branched U signals to
generate simultaneous linear equations, and solves the simultaneous
linear equations for receiving transmission signals with no
interference.
[0162] That is,
(1) The transmitting side makes the signals pass through P
different virtual channels created on the transmitting side, adds
them up, and transmits the signals via the actual transmission
channels, and the receiving side makes the reception signals pass
through U (U.gtoreq.P) different virtual channels. (2) Because
there are P virtual channels on the transmitting side and U virtual
channels on the receiving side, the signals pass through the
virtual channels in PU different ways. The receiving side detects
all those PU virtual channel characteristics.
[0163] Out of the characteristics of all PU virtual channels, the
receiving side detects at least P.sup.2 virtual channels.
(3) Using the PU channel characteristics obtained by the receiving
side and the outputs that are the data transmitted from the
transmitting side and passing through the separate receiving-side
virtual channels, the receiving side generates simultaneous linear
equations and solves the simultaneous linear equations for
receiving the transmission signals with no interference.
[0164] As the method for inputting signals to U (U.gtoreq.P)
separate virtual channels on the receiving side, there are two
methods; one is an oversampling method and the other is a
convolution method for performing the convolution between the
signals and U pieces of virtual channel generation data in the same
way the convolution is performed on the transmitting side.
[0165] FIG. 12 is a diagram showing the oversampling method, and
FIG. 44 is a diagram showing the method for performing the
convolution between the signal and virtual channel generation
data.
Receiving Device (1)
[0166] With reference to FIG. 12, the following describes a
receiving device that performs oversampling for the signal
transmitted from the transmitting device shown in FIG. 9. Note that
FIG. 12 shows a general case, not a case when R(P)=3.
[0167] The receiving device shown in FIG. 12 comprises an antenna
21, a receiving unit 22 that converts the reception signal,
detected by the antenna 21, to a baseband signal, an oversampling
unit 28 that performs the U-fold oversampling of the reception
signal that has been converted to the baseband signal by the
receiving unit 22, a signal separation unit 29 that is configured
by matched filters each of which matches to the Kronecker product
of each of N row vectors, that is, vector f.sub.N,0, vector
f.sub.N,1 . . . vector f.sub.N,N-1, and the vector I.sub.M, for
each sampling sequence that is output from the oversampling unit 28
and will be described later, a channel characteristics detection
unit 25 that detects the channel characteristics on the time axis
of all transmission lines from the transmitting side to the
receiving side including the transmission characteristics of the
transmitting device side, the transmission characteristics of the
propagation space, and the transmission characteristics of the
receiving device side (detects the channel characteristics on the
time axis of all combinations of the virtual transmission channels
and virtual reception antennas which will be described later), a
simultaneous equation generation unit 26, and a decoding unit
27.
Oversampling
[0168] The oversampling unit 28 performs the U-fold oversampling of
the reception signal converted to the baseband signal by the
receiving unit 22.
[0169] The following schematically describes the oversampling of
the signal P(1, -1, 1, 1) with reference to FIG. 13. Note that a
signal to be oversampled by the oversampling unit 28 is not a
signal composed of clear "0" and "1", such as the one shown in FIG.
13, but an unclear signal including line noises and thermal noises
or a signal including leak signals from other channels.
[0170] When the pitch interval of the signal P is .tau. (pitch
frequency 1/.tau.) as shown in FIG. 13(A), oversampling at the
frequency four time higher than the pitch frequency (interval of
.tau./4) changes the signal A(1, -1, 1, 1) to the signal B(1, 1, 1,
1, -1, -1, -1, -1, 1, 1, 1, 1, 1, 1, 1, 1) as shown in FIG.
13(B).
[0171] If the oversampling timing is fixed, the oversampling
interval need not be the same.
[0172] That is, this is described as follows using # that will be
used later. For any #, the sampling interval in the same # is fixed
(.tau./4 in the figure), but the interval between #'s need not be
fixed.
[0173] Oversampling may be performed, not only after the signal is
converted to a baseband signal, but when the signal is a
high-frequency or intermediate-frequency signal.
[0174] As shown in FIG. 14, the signal (a.sub.0, a.sub.1, . . . ,
a.sub.(M-1)) is received and oversampled. In this figure, the
oversampling #0 signal for the signal a.sub.0 is a.sub.0-0, the
oversampling #1 signal for the signal a.sub.0 is a.sub.0-1, . . . ,
the oversampling #(U-1) signal for the signal a.sub.0 is
a.sub.0-(U-1), the oversampling #0 signal for the signal a.sub.1 is
a.sub.1-0, the oversampling #1 signal for the signal a.sub.1 is
a.sub.1-1, . . . , the oversampling #(U-1) signal for the signal
a.sub.1 is a.sub.1-(U-1), . . . , the oversampling #0 signal for
the signal a.sub.(M-1) is a.sub.(M-1)-0, the oversampling #1 signal
for the signal a.sub.(M-1) is a.sub.(M-1)-1, . . . , and the
oversampling #(U-1) signal for the signal a.sub.(M-1) is
a.sub.(M-1)-(U-1).
[0175] FIG. 15 shows the signal sequences at the sampling
point.
[0176] Sampling #0 sequence a.sub.0-0a.sub.1-0 . . .
a.sub.(M-1)-0
[0177] Sampling #1 sequence a.sub.0-1a.sub.1-1 . . .
a.sub.(M-1)-1
[0178] . . .
[0179] Sampling #U-1 sequence a.sub.0-(U-1)a.sub.1-(U-1) . . .
a.sub.(M-1)-(U-1)
[0180] This indicates that, for each sampling sequence, there is a
signal sequence corresponding to the transmission signal; in other
words, it can be said that there is a virtual channel for each
sampling sequence. This virtual channel, which is generated on the
receiving side, is thought of as a virtual reception antenna.
[0181] Note that, because the pilot signals use different row
vectors for each virtual channel, all pilot signals can be received
by the receiving side with no interference. On the other hand,
because transmission data does not use different row vectors for
each virtual channel, interference occurs among multiple pieces of
transmission data that use the same row vector.
[0182] In the present invention, the receiving side performs
oversampling, generates the simultaneous linear equations for
decoding the transmission data, and solves the simultaneous linear
equations to eliminate the effect of channel characteristics of the
lines for estimating the transmission data.
[0183] In this case, the oversampling U satisfies the following
relation.
U.gtoreq.R
Signal Separation
[0184] The signal separation unit 29 inputs each sampling sequence,
which is output by the oversampling unit 28, into the matched
filters each of which matches to the Kronecker product of each of
the N row vectors, that is, vector f.sub.N,0, vector f.sub.N,1 . .
. vector f.sub.N,N-1, and the vector I.sub.M. The signal separation
unit 29 separates the signal for each matched filter that matches
to the Kronecker product of each of the N row vectors, that is,
vector f.sub.N,0, vector f.sub.N,1 . . . vector f.sub.N,N-1, and
the vector I.sub.M.
[0185] The resulting separated signals are the P pilot signal and
N-P pieces of transmission data for each sampling sequence.
[0186] Next, the following describes the separation of the pilot
signal in a sampling sequence #i (0.ltoreq.i.ltoreq.U-1).
[0187] In the virtual channel 0 (virtual transmission channel 0,
virtual transmission antenna 0) transmission signal KS0, the pilot
signal X.sub.0
[Mathematical Expression 14]
[0188] is inserted as Vector f.sub.N,0Pilot signal X.sub.0.
[0189] Therefore, the pilot signal X.sub.0 may be produced by
inputting the sampling sequence into the matched filter of the
Kronecker product of vector f.sub.N,0 and vector I.sub.M.
[0190] Similarly, the signal separation unit 29 inputs the signals,
received via virtual channel 1 (virtual transmission channel 1,
virtual transmission antenna 1) and via virtual channel 2 (virtual
transmission channel 2, virtual transmission antenna 2), into the
matched filter of the Kronecker product of vector f.sub.N,1 and
vector I.sub.M and into the matched filter of the Kronecker product
of vector f.sub.N,2 and vector I.sub.M to produce the pilot signal
X.sub.1 and the pilot signal X.sub.2W of virtual channel 1 and
virtual channel 2.
[0191] The N-P pieces of transmission data can be produced in the
same way as the pilot signal is extracted.
[0192] That is, the signal separation unit 29 inputs the signal,
transmitted via virtual channel 0, into the matched filter of the
Kronecker product of each of the N-P (in this case, P=3) row
vectors f.sub.N,P-f.sub.N,N-1 and the vector I.sub.M, to produce
N-P pieces of transmission data X.sub.0,P(x.sub.0,P,0, x.sub.0,P,1,
. . . , x.sub.0,P,(M-1)) . . . X.sub.0,N-1(x.sub.0,(N-1),0,
x.sub.0,(N-1),1, . . . , x.sub.0,(N-1),(M-1)).
[0193] Similarly, the signal separation unit 29 produces (N-P)
pieces of transmission data, transmitted via virtual channel 1 and
virtual channel 2, by inputting the signal, transmitted via virtual
channel 1 and virtual channel 2, into the matched filters of the
Kronecker product of each of N-P row vectors f.sub.N,P-f.sub.N,N-1
and the vector I.sub.M.
[0194] Note that, because the pilot signals use different row
vectors for each virtual channel (each virtual transmission
channel, each virtual transmission antenna), all pilot signals can
be received by the receiving side with no interference. On the
other hand, because transmission data does not use different row
vectors for each virtual channel, interference will occur among
multiple pieces of transmission data that use the same row vector
if no action is taken.
[0195] In the present invention, the transmitting side transmits
different pilot signals onto virtual channels (virtual transmission
channels, virtual transmission antennas), one for each. The
receiving side, which receives those pilot signals, can detect the
channel characteristics including the channel characteristics of
all virtual channels. Those channel characteristics allow the
receiving side to detect transmission data with no
interference.
Detection of Channel Characteristics
[0196] In the present invention, one pilot signal is inserted for
each virtual channel (virtual transmission antenna) as described
below.
[Mathematical Expression 15]
[0197] The pilot signal X.sub.0 is inserted into the virtual
channel 0 transmission signal KS0 as Vector f.sub.N,0Pilot signal
X.sub.0, and no signal is related to f.sub.N,1 and the vector
f.sub.N,2.
[0198] Similarly, the pilot signal X.sub.1 is inserted into the
virtual channel 1 transmission signal KS1 as Vector f.sub.N,1Pilot
signal X.sub.1, and no signal is related to the vector f.sub.N,0
and the vector f.sub.N,2.
[0199] Similarly, the pilot signal X.sub.2 is inserted into the
virtual channel 2 transmission signal KS2 as Vector f.sub.N,2Pilot
signal X.sub.2, and no signal is related to the vector f.sub.N,0
and the vector f.sub.N,1.
[0200] Therefore, the channel characteristics of #j virtual channel
(virtual transmission antenna) can be obtained by detecting #j
(0.ltoreq.j.ltoreq.N-1) pilot signal.
[0201] The present invention provides each sampling sequence
with
[0202] a matched filter for the Kronecker product of vector f.sub.0
and vector I.sub.M
[0203] a matched filter for the Kronecker product of vector f.sub.1
and vector I.sub.M
.cndot.
.cndot.
[0206] and a matched filter for the Kronecker product of vector
f.sub.p-1 and vector I.sub.M.
[0207] In other words,
[0208] each of the virtual reception antennas #0-#(U-1) has
[0209] a matched filter for the Kronecker product of vector f.sub.0
and vector I.sub.M
[0210] a matched filter for the Kronecker product of vector f.sub.1
and vector I.sub.M
.cndot.
.cndot.
[0213] and a matched filter for the Kronecker product of vector
f.sub.p-1 and vector I.sub.M.
[0214] Note that the output, generated by inputting the signal,
received from the virtual reception antenna #i
(0.ltoreq.i.ltoreq.U-1), into the matched filter for the Kronecker
product of vector f.sub.j (0.ltoreq.j.ltoreq.P-1) and vector
I.sub.M is the characteristics of the virtual channel from the
virtual transmission antenna j to the virtual reception antenna
i.
[0215] Therefore, at the virtual reception antenna #i
(0.ltoreq.i.ltoreq.U-1) (that is, oversampling #i sequence), the
virtual channel characteristics between all virtual transmission
antennas and virtual reception antenna #i can be obtained.
[0216] The channel characteristics detection unit 25 performs this
processing at all virtual reception antennas to obtain the virtual
channel characteristics between all virtual transmission antennas
and all virtual reception antennas.
[0217] Because the channel characteristics are detected on the time
axis in the present invention, the channel characteristics
detection unit 25 can detect the channel characteristics on the
time axis on all transmission lines from the transmitting side to
the receiving side including the transmission characteristics of
the transmitting device side, the transmission characteristics of
the propagation space, and the transmission characteristics of the
receiving device side.
[0218] In the present invention, the channel characteristics on the
time axis are detected and the time response at digital signal
transmission time is detected.
[0219] The channel characteristics on the time axis are the channel
characteristics of all transmission lines from the transmitting
side to the receiving side, which include the transmission
characteristics of the transmitting device side, the transmission
characteristics of the propagation space, and the transmission
characteristics of the receiving device side, including those of
multipath responses.
[0220] The channel characteristics on the time axis are represented
in the form similar to that of the multipath characteristics.
Generation of Simultaneous Equations
[0221] The simultaneous equation generation unit 26 generates
simultaneous equations, such as those shown in Expression (17),
based on the reception signal generated via oversampling and on the
channel characteristics on the time axis detected by the channel
characteristics detection unit 25.
[0222] In practice, the simultaneous equation generation unit 26
generates T.sub.0, T.sub.1, and T.sub.2 by adding up the
transmission data related to the corresponding row vectors as shown
in FIG. 16 and generates the simultaneous equations for each of
T.sub.0, T.sub.1, and T.sub.2.
Decoding
[0223] The simultaneous equation generation unit 26 generates the
simultaneous equations, such as those in Expression (17), based on
the reception signals of the three virtual channels separated by
the channel separation unit 29 and on the channel characteristics
on the time axis detected by the channel characteristics detection
unit 25.
[0224] The decoding unit 27 solves the simultaneous equations
generated by the simultaneous equation generation unit 26. Because
the channel characteristics on the time axis of all transmission
lines from the transmitting side to the receiving side, including
the transmission characteristics of the transmitting device side,
the transmission characteristics of the propagation space, and the
transmission characteristics of the receiving device side, are
reflected on the simultaneous equations generated by the
simultaneous equation generation unit 26, the solution can decode
the transmission data not affected by the transmission
characteristics of the transmitting device side, the transmission
characteristics of the propagation space, and the transmission
characteristics of the receiving device side.
[0225] In other words, solving the simultaneous equations in
Expression (17) gives the signals from which the effect of the
channel characteristics on the time axis of the transmission lines
is removed.
[0226] As described above, the virtual channels, generated from the
virtual channel generation data, and the virtual channels,
generated for each oversampling sequence, are obtained. The former
are virtual channels generated on the transmitting side, the latter
are virtual channels generated on the receiving side, and those
virtual channels are generated independently. Thus, it can be said
that the virtual transmission antennas are generated by the former
and that the virtual reception antennas are generated by the
latter.
[0227] When three virtual channels are generated from the virtual
channel generation data (P=3) and U virtual channels are generated
for each oversampling sequence, the number of channels is 3U.
[0228] Therefore, for the 3U channels, the channel characteristics
detection unit 25 detects the channel characteristics on the time
axis of all transmission lines from the transmitting side to the
receiving side, including the transmission characteristics of the
transmitting device side, the transmission characteristics of the
propagation space, and the transmission characteristics of the
receiving device side.
Receiving Device (2)
Method for Generating Virtual Channels by Convolution Performed on
the Receiving Side
[0229] Next, with reference to FIG. 44, the following describes a
method for generating virtual channels on the receiving side in
which reception signals are convoluted with virtual channel
generation data.
[0230] A receiving device in FIG. 44 comprises an antenna 21, a
receiving unit 22 that converts the reception signal, detected by
the antenna 21, to a baseband signal, a receiving side virtual
channel convolution unit 38 that performs the convolution between
the reception signals, converted to a baseband signal by the
receiving unit 22, and the U.sub.1 pieces of virtual channel
generation data, a signal separation unit 29 that is configured by
matched filters each of which matches to the Kronecker product of
each of N row vectors, that is, vector f.sub.N,0, vector f.sub.N,1
. . . vector f.sub.N,N-1, and the vector I.sub.M, for each of the
U.sub.1 virtual channels output from the receiving side virtual
channel convolution unit 38, a channel characteristics detection
unit 25, a simultaneous equation generation unit 26, and a decoding
unit 27.
[0231] In FIG. 44, the components other than the receiving side
virtual channel convolution unit 38, that is, the antenna 21,
receiving unit 22, signal separation unit 29, channel
characteristics detection unit 25, simultaneous equation generation
unit 26, and decoding unit 27 are the same as those in FIG. 12.
[0232] The receiving side virtual channel convolution unit 38
comprises a virtual channel generation data storage unit 381 (need
not always comprise the virtual channel generation data storage
unit 381) and performs the convolution between the reception
signal, converted to a baseband signal by the receiving unit 22,
and one of virtual channel generation data stored in the virtual
channel generation data storage unit 381.
[0233] When the reception signal is RS and the virtual channel
generation data is E.sub.1-E.sub.U1, the receiving side virtual
channel convolution unit 38 performs the convolution between the
reception signal RS and the virtual channel generation data
E.sub.1-E.sub.U1 and separately outputs
[0234] signal #1 generated by the convolution between the reception
signal RS and virtual channel generation data E.sub.1,
[0235] signal #2 generated by the convolution between the reception
signal RS and virtual channel generation data E.sub.12,
.cndot.
.cndot.
[0238] and signal #U.sub.1 generated by the convolution between the
reception signal RS and virtual channel generation data
E.sub.U1.
[0239] When the U.sub.1 outputs are received from the receiving
side virtual channel convolution unit 38, the signal separation
circuit separates the signals via the matched filters each of which
matches to the Kronecker product of each of the N row vectors, that
is, vector f.sub.N,0, vector f.sub.N,1, . . . , and vector
f.sub.N,N-1, and vector I.sub.M.
[0240] The channel characteristics detection unit 25, simultaneous
equation generation unit 26, and decoding unit 27 are the same as
those in FIG. 12 and so the description is omitted here.
Number of Antennas
[0241] In the above description, there is one transmission antenna
and one reception antenna, and there are R virtual channel
transmission antennas and U virtual channel reception antennas.
[0242] Assume that the number of virtual channels is R, the number
of virtual channel transmission antennas is T, the number of
virtual channel reception antennas is V, the number of actual
transmission antennas is TA, and the number of actual reception
antennas is RA.
[0243] In a typical system,
[0244] Number of virtual channels R=Number of virtual channel
reception antennas U and
[0245] Number of actual transmission antennas TA=Number of actual
reception antennas RA=1.
[0246] The present invention is not limited to the case described
above.
[0247] For example, there are the following four cases, case 1-case
4.
[0248] (1) Case 1
[0249] Transmitting side: The number of actual transmission
antennas is R.
[0250] Receiving side: The number of actual reception antennas is
1, and the number of virtual channel reception antennas is R.
[0251] (2) Case 2
[0252] Transmitting side: The number of actual transmission
antennas is 1, and the number of virtual channel transmission
antennas is R.
[0253] Receiving side: The number of actual reception antennas is
R.
[0254] (3) Case 3
[0255] Transmitting side: The number of actual transmission
antennas is R.
[0256] Receiving side: The number of actual reception antennas is
R.
[0257] (4) Case 4
[0258] Transmitting side: The number of actual transmission
antennas is TA.
[0259] The number of virtual channel transmission antennas is
(R-TA).
[0260] Receiving side: The number of actual reception antennas is
RA.
[0261] The number of virtual channel reception antennas is
(R-RA).
[0262] In the embodiment described above, the number of virtual
transmission antennas is "3".
[0263] FIG. 17 is a diagram showing the case in which the number of
actual transmission antennas R on the transmitting side is equal to
the number of virtual channels R.
[0264] A transmitting device in FIG. 17 comprises a virtual channel
0 transmission signal creation unit 11, a virtual channel 1
transmission signal creation unit 12, a virtual channel 2
transmission signal creation unit 13, virtual channel generation
data 151, convolution units 152-154, a transmitting unit 171, and
an antenna 181.
[0265] The convolution unit 152 performs the convolution between
the virtual channel 0 transmission signal KS0 and virtual channel
generation data D2, the convolution unit 153 performs the
convolution between the virtual channel 1 transmission signal KS1
and virtual channel generation data D1, and the convolution unit
154 performs the convolution between the virtual channel 2
transmission signal KS2 and virtual channel generation data D0.
[0266] The transmitting unit 171 converts the signals (virtual
channel 0 transmission signal, virtual channel 1 transmission
signal, virtual channel 2 transmission signal), which are received
from the convolution units 152-154, to the high frequency signals
and transmits them via separate antennas 181, 182, and 183.
[0267] FIG. 18 is a diagram showing the case in which the number of
actual transmission antennas R on the transmitting side is smaller
than the number of virtual channels R and, when the number of
actual transmission antennas is TA, the number of virtual channel
transmission antennas is R-TA.
[0268] The transmitting unit may be provided, one for each of the
antennas 181-183. In this case, three users may use separate
antennas.
[0269] The transmitting device in FIG. 18 comprises a virtual
channel 0 transmission signal creation unit 11, a virtual channel 1
transmission signal creation unit 12, a virtual channel 2
transmission signal creation unit 13, virtual channel generation
data 151, convolution units 152-154, an addition unit 1551, a
transmitting unit 172, and antennas 184 and 185.
[0270] The convolution unit 152 performs the convolution between
the virtual channel 0 transmission signal KS0 and the virtual
channel generation data D2, the convolution unit 153 performs the
convolution between the virtual channel 1 transmission signal KS1
and the virtual channel generation data D1, and the convolution
unit 154 performs the convolution between the virtual channel 2
transmission signal KS2 and the virtual channel generation data
D0.
[0271] The addition unit 1551 adds up the signals from the
convolution units 153 and 154 and outputs the result to the
transmitting unit 17.
[0272] The output of the convolution unit 152 is output directly to
the transmitting unit 17.
[0273] The transmitting unit 171 converts the signal (virtual
channel 0 transmission signal) from the convolution unit 152 and
the signal (virtual channel 1 transmission signal, virtual channel
2 transmission signal) from the addition unit 151 to the high
frequency signal and transmits the converted high frequency signal
via the antennas.
[0274] The virtual channel 0 transmission signal is transmitted
from the antenna 184, and the virtual channel 1 transmission signal
and the virtual channel 2 transmission signal are transmitted from
the antenna 185.
[0275] FIG. 19 is a diagram showing the case in which the number of
actual transmission antennas R on the receiving side is equal to
the number of virtual channels R.
[0276] The transmitting units may be provided for the antennas
181-183, one for each. In this case, two users may use separate
antennas.
[0277] FIG. 19 is a diagram showing that the signal transmitted
from the transmitting device in FIG. 9, FIG. 17, or FIG. 18 is
received via the antennas 211-213 and receiving unit 221-223.
[0278] The receiving device in FIG. 19 comprises the antennas
211-213, an oversampling unit 28 that oversamples the reception
signals detected by the antennas 211-213, a signal separation unit
29 that is configured by matched filters matching to the N row
vectors, that is, vector f.sub.N,0, vector f.sub.N,1 . . . vector
f.sub.N,N-1, for separating the output of the oversampling unit 28
into the signal for each oversampling sequence, a channel
characteristics detection unit 25 that detects the channel
characteristics on the time axis on all transmission lines from the
transmitting side to the receiving side including the transmission
characteristics of the transmitting device side, the transmission
characteristics of the propagation space, and the transmission
characteristics of the receiving device side, a simultaneous
equation generation unit 26, and a decoding unit 27.
[0279] The oversampling unit 28 may also provided for the receiving
units 221-223, one for each.
[0280] When the receiving units 221-223 are used by different
users, the receiving device is configured to provide the
oversampling unit for each user.
[0281] FIG. 20 is a diagram showing the case in which the number of
actual transmission antennas R on the receiving side is smaller
than the number of virtual channels R.
[0282] FIG. 20 is a diagram showing that the signal transmitted
from the transmitting device in FIG. 9, FIG. 17, or FIG. 18 is
received via antennas 214 and 215 and receiving units 224 and
225.
[0283] The receiving device in FIG. 20 comprises the antennas 214
and 215, receiving units 224 and 225 that convert the reception
signal detected by the antennas 214 and 215 to the baseband signal,
an oversampling unit 28, a signal separation unit 29 that is
configured by matched filters matching to the N row vectors, that
is, vector f.sub.N,0, vector f.sub.N,1 . . . vector f.sub.N,N-1,
for separating the signal into the signal for each oversampling
sequence, a channel characteristics detection unit 25 that detects
the channel characteristics on the time axis on all transmission
lines from the transmitting side to the receiving side including
the transmission characteristics of the transmitting device side,
the transmission characteristics of the propagation space, and the
transmission characteristics of the receiving device side, a
simultaneous equation generation unit 26, and a decoding unit
27.
[0284] According to the above description, case 1 described above
is the case in which the transmitting side is the device shown in
FIG. 17 and the receiving side is the device shown in FIG. 12, case
2 described above is the case in which the transmitting side is the
device shown in FIG. 9 and the receiving side is the device shown
in FIG. 19, case 3 described above is the case in which the
transmitting side is the device shown in FIG. 17 and the receiving
side is the device shown in FIG. 21, and case 4 described above is
the case in which the transmitting side is the device shown in FIG.
18 and the receiving side is the device shown in FIG. 20.
[0285] If the number of virtual channel transmission antennas T=the
number of virtual channel reception antennas U, the receiving side
generates one simultaneous equation similar to Expression (17).
[0286] However, if the number of virtual channel transmission
antennas T<the number of virtual channel reception antennas V,
the receiving side can select T antennas from all virtual antennas
V and generate the simultaneous equations.
[0287] The number of selections is given by the following
expression.
.sub.VC.sub.T [Mathematical Expression 16]
[0288] As a result, the receiving side can generate multiple
simultaneous equations similar to Expression (17).
[0289] In this case, because multiple estimation results are
obtained for the same transmission data, the decision by majority
or some other method may be used to estimate probable transmission
data to reduce the bit error rate.
[0290] Even in a case other than when the number of virtual channel
transmission antennas T<the number of virtual channel reception
antennas V, multiple simultaneous equations may be generated by not
transmitting information onto one or more virtual antennas on the
transmitting side as in the case when the number of virtual channel
transmission antennas T<the number of virtual channel reception
antennas V.
Pilot Signal
[Mathematical Expression 17]
[0291] When the pilot signal is X.sub.K(x.sub.k0, x.sub.K1,
x.sub.K2, . . . , x.sub.K(M-1)) and the pilot signal is the Kth row
vector of the Nth order DFT matrix (W.sub.N.sup.0, W.sub.N.sup.k,
W.sub.N.sup.2k . . . W.sub.N.sup.(N-1)k), the pilot signal is as
shown below and there is a need to consider the effect of the
multiplication of re.sup.j.theta.1 such as the one shown in FIG.
21.
S k = f k X k = ( W N 0 x k 0 , W N 0 x k 1 , W N 0 x k 2 , , W N 0
x k ( M - 1 ) , W N k x k 0 , W N 0 x k 1 , W N k x k 2 , , W N k x
k ( M - 1 ) , W N ( N - 1 ) k x k 0 , W N ( N - 1 ) k x k 1 , W N (
N - 1 ) k x k 2 , , W N ( N - 1 ) k x k ( M - 1 ) , )
##EQU00005##
[Mathematical Expression 18]
[0292] To avoid this problem, the pilot signal X.sub.K'(x.sub.k0,
W.sub.MN.sup.1x.sub.k1, W.sub.MN.sup.2x.sub.k2, . . . ,
W.sub.MN.sup.(M-1)x.sub.k(M-1)) is used instead of the pilot signal
X.sub.K(x.sub.K0, x.sub.K1, x.sub.K2, . . . , x.sub.K(M-1)). The
result is as shown below and, in this case, there is no need to
consider the effect of the multiplication of re.sup.j.theta.1 such
as the one shown in FIG. 21.
S k ' = ( f k X k , = ( W MN 0 x k 0 , W MN 1 x k 1 , W MN 2 x k 2
, , W MN ( M - 1 ) x k ( M - 1 ) , W MN M x k 0 , W MN M + 1 x k 1
, W MN M + 2 x k 2 , , W MN ( 2 M - 1 ) x k ( M - 1 ) , W MN M ( N
- 1 ) x k 0 , W MN M ( N - 1 ) + 1 x k 1 , W MN M ( N - 1 ) + 2 x k
2 , , W MN MN - 1 x k ( M - 1 ) , ) ##EQU00006##
[Mathematical Expression 19]
[0293] In addition, the pilot signal X.sub.K''(x.sub.k0,
W.sub.MN.sup.ux.sub.k1, W.sub.MN.sup.2ux.sub.k2, . . . ,
W.sub.MN.sup.u(M-1)x.sub.k(M-1)) may be used instead of the pilot
signal X.sub.K'(x.sub.k0, W.sub.MN.sup.1x.sub.k1,
W.sub.MN.sup.2x.sub.k2, . . . ,
W.sub.MN.sup.(M-1)x.sub.k(M-1)).
[0294] Note that such a problem is not generated because all the
elements of the row vector f.sub.0(W.sub.N.sup.0, W.sub.N.sup.0, .
. . , W.sub.N.sup.0) are equal to 1/ {square root over ( )}N.
[0295] Therefore, if there is no need to make the channel
estimation for the virtual channels with the use of the pilot
signal for each transmission, the channel estimation can be made
without using the pilot signal X.sub.K' and the pilot signal
X.sub.K'' by allocating the row vector f.sub.0 alternately for the
virtual channels.
[0296] In this case, for each virtual channel (virtual transmission
antenna), the channel characteristics of the virtual channels,
detected by receiving the pilot signal, are used as the channel
characteristics of the virtual channels when the pilot signal is
not received.
[0297] FIG. 22 is a diagram showing the case where the virtual
channel (virtual transmission antenna) R=3.
[0298] In phase 1, the pilot signal f.sub.0 is transmitted over
virtual channel 0, and the pilot signal is not transmitted over
virtual channels 1 and 2.
[0299] In phase 2, the pilot signal f.sub.0 is transmitted over
virtual channel 1, and the pilot signal is not transmitted over
virtual channels 0 and 2.
[0300] At this time, the channel characteristics detected in phase
1 are used as the channel characteristics of virtual channel 0.
[0301] In phase 3, the pilot signal f.sub.0 is transmitted over
virtual channel 2, and the pilot signal is not transmitted over
virtual channels 0 and 1.
[0302] At this time, the channel characteristics detected in phase
1 are used as the channel characteristics of virtual channel 0, and
the channel characteristics detected in phase 2 are used as the
channel characteristics of virtual channel 1.
[0303] Phase 3 is followed by phase 1, and phase 1, phase 2, and
phase 3 are repeated in a circular fashion.
[0304] In phase 1 that follows, the channel characteristics
detected in this phase 1 are used as the channel characteristics of
virtual channel 0, the channel characteristics detected in the
previous phase 2 are used as the channel characteristics of virtual
channel 1, and the channel characteristics detected in the previous
phase 3 are used as the channel characteristics of virtual channel
2.
[0305] On the transmitting side, it is only required that P pilot
sequences, each having a spectrum in the shape of the teeth of a
comb, do not interfere with each other and with the data
signal.
[0306] It is only required for the data signal to have a spectrum
in the shape of the teeth of a comb so that one piece of data
transmitted over a virtual channel does not interfere with another
piece of data transmitted over the same virtual channel. That is, a
non-DFT row vector may be used to generate signals from data.
[0307] On the receiving side, it is only required that a sequence
of matched filters, which have a synchronous spectrum in the shape
of the teeth of a comb, are used to receive each of the P pilot
signals.
[0308] A sequence of any matched filters, which have a periodic
spectrum in the shape of the teeth of a comb, may be used to
receive the data signal, and a non-DFT row vector may also be
used.
Transmission Data
[0309] In the signal configuration in FIG. 5 to FIG. 7, N-P pieces
of transmission data X.sub.0,P(x.sub.0,P,0, x.sub.0,P,1, . . . ,
x.sub.0,P,(M-1)) . . . X.sub.0,N-1(x.sub.0,(N-1),0,
x.sub.0,(N-1),1, . . . , x.sub.0,(N-1),(M-1)) are transmitted over
virtual channel 0, N-P pieces of transmission data
X.sub.1,P(x.sub.1,P,0, x.sub.1,P,1, . . . , x.sub.1,P,(M-1)) . . .
X.sub.1,N-1(x.sub.1,(N-1),0, x.sub.1,(N-1),1, . . . ,
x.sub.1,(N-1),(M-1) are transmitted over virtual channel 1, and N-P
pieces of transmission data X.sub.2,P(x.sub.2,P,0, x.sub.2,P,1, . .
. , x.sub.2,P,(M-1)) . . . X.sub.2,N-1(x.sub.2,(N-1),0,
x.sub.2,(N-1),1, . . . , x.sub.2,(N-1),(M-1) are transmitted over
virtual channel 2.
[0310] Data X.sub.0,P . . . X.sub.0,N-1 is transmitted over virtual
channel 0, data X.sub.1,P . . . X.sub.1,N-1 is transmitted over
virtual channel 1, and data X.sub.2,P . . . X.sub.2,N-1 is
transmitted over virtual channel 2.
[0311] Data transmitted over the virtual channels may be the same
data or different data.
[0312] This applies also to case 1, case 2, case 3, and case 4
given above.
[0313] In this case, actual antennas may be used by different
users, one for each user.
[0314] In such an environment, there may be multiple transmitting
users and one receiving user or there may be one transmitting user
and multiple receiving users.
[0315] When N=1024 and the vectors for the pilot signals (pilot
sequences) corresponding to all virtual channels of multiple users
are reserved, the remaining data vectors (data sequences) may be
shared by multiple virtual channels (even if multiple users are
allocated).
[0316] That is, because only the pilot signals require extra
bandwidths even if there are many users, the frequency usage
efficiency is further increased.
[0317] Note that not only the pilot signals but also some other
signals, such those for adjusting bandwidth distributions, vary
according to the users. However, those adjusting signals may be
shared on a user basis.
[0318] This ability is provided because different data is
transmitted over virtual channels. MIMO-OFDM, in which the same
signal (or information) is transmitted from all antennas, does not
provide such a usage method.
[0319] If transmitting or receiving users use one or more actual
antennas to transmit or receive data over one or more virtual
channels and if
(A) there are many transmitting users, that is, there are multiple
actual antennas and (B) the receiving side is a base station
(single user) and there is one or more actual antennas,
[0320] the present invention, in which virtual channel generation
data is used to generate virtual channels (virtual transmission
channels, virtual transmission antennas), allows the transmitting
side to control power to avoid the near-far problem, thus reducing
the generation of reception noises.
[0321] The power control data is generated on the receiving side
based on the received noises and is notified to the transmitting
side.
[0322] Next, the following describes "technical basis of the
invention--theory of OSDM" and "multiple virtual antenna OSDM
system" corresponding to the embodiments of the present
invention.
Technical Background of the Invention
Theory of OSDM
[0323] The following describes the theory of OSDM that is the
technical basis of the present invention.
1. (Chapter 1) Foreword
[0324] Recently, social needs for telecommunication become more
widespread and diversified as the information society evolves. In
particular, mobile information communication, which uses the
wireless technology to allow the user to communicate while moving
around, is an important element indispensable for the social
infrastructure.
[0325] The OFDM system, which is recognized as a next generation
communication technology, is one of the technologies that attract
attention in various fields because of its high frequency usage
efficiency and high anti-multipath feature. On the other hand, a
problem is pointed out that, because independently-modulated
carriers are superposed, the Peak to Average Power Ratio (PAPR)
becomes high. Like the OFDM system, the OSDM system is a
communication system designed to improve the frequency usage
efficiency for providing a drastically higher communication path
capacity than that of the other communication systems and, at the
same time, provides the user with the real-time acquisition of the
communication path environment that has been impossible in the
conventional systems. It is reported that the PAPR is almost flat
as compared with the OFDM system. Applying this feature to a
multi-antenna communication system allows the communication path
capacity to be almost proportionally increased to the number of
transmission/reception antennas. The following evaluates the
performance of the OSDM system, from the basics to a multi-antenna
OSDM system that is an extended version, based on the simulation
results while comparing the OSDM system with the OFDM system as
necessary.
[0326] The description is composed of six chapters. Chapter 2
introduces the basic theory of the OSDM system, and Chapter 3 that
follows evaluates the performance trough simulation. Chapter 4
proposes the theory of a multi-antenna OSDM system that is an
extended version of the OSDM system, and Chapter 5 evaluates the
performance of the multi-antenna OSDM system through simulation.
Chapter 6 summarizes the information collected from those
results.
2. Basic Theory
[0327] This chapter introduces the theory of the OSDM system in
four sections. First, Section 1 introduces the process of forming a
transmission signal from data with focus on the transmission
system. Next, Section 2 introduces the process of acquiring the
communication path environment from the reception signal and
estimating data from the acquired communication path environment
with focus on the receiving system. Finally, Section 3 describes
the features of the OSDM system by comparing them with those of the
OFDM system.
2.1 Transmission System (Section 1 of Chapter 2)
[0328] The transmission system is configured by the processes shown
in FIG. 23. The following describes the details.
The data vectors x.sub.0, x.sub.1, . . . , x.sub.N-1, each having
the length M, are defined as follows.
x 0 = ( x 00 , x 01 , , x 0 ( M - 1 ) ) x 1 = ( x 10 , x 11 , , x
01 ( M - 1 ) ) x N - 1 = ( x ( N - 1 ) 0 , x ( N - 1 ) 1 , , x 0 ,
( N - 1 ) ( M - 1 ) ) ( 25 ) ##EQU00007##
When W.sub.N.ident.exp(2.pi. {square root over ( )}-1)/N, the Nth
order inverse-DFT matrix F.sup.-1 and its row vectors, that is, row
vector f.sub.N,0, row vector f.sub.N,1 . . . row vector f.sub.N,N-1
(hereinafter called "row vector f.sub.0, row vector f.sub.1 . . .
row vector f.sub.N-1" or "f.sub.0, f.sub.1 . . . f.sub.N-1"), are
defined as shown in FIG. 4.
[0329] The Kronecker product of the vector f.sub.i and x.sub.i is
X.sub.i.
[0330] That is,
[ Mathematical expression 20 ] X i = f i dataX i ( i = 0 , 1 , N -
1 ) ( 26 ) Sum of X i S sum = ( S 0 , S 1 , , S MN - 1 ) = i = 0 N
- 1 X i ( 27 ) ##EQU00008##
[0331] The signal generated by adding the cyclic prefix having the
length L-1 to the signal shown above
S=(S.sub.MN-L+1, . . . , S.sub.MN-1,S.sub.0,S.sub.1, . . . ,
S.sub.MN-1) (28)
is the signal actually transmitted to the communication paths.
2.2 Reception System
[0332] The reception system is composed of the processes shown in
FIG. 24. The following describes the details.
[0333] Let h.sub.0, h.sub.1, . . . , h.sub.L-1, be the impulse
responses on the communication paths. The reception signal
.sup..about.R is represented by the following expression using
MN.times.MN right cyclic shift matrix T.
[ Mathematical expression 21 ] R _ = i = 0 L - 1 h i ST i ( 29 )
##EQU00009##
[0334] Let R be the signal generated by removing the cyclic prefix
from {tilde over (R)}. Then, there is the following relation
between R and the data vectors x.sub.0, x.sub.1, . . . ,
x.sub.N-1.
Y.sup.def=RW(x.sub.0,x.sub.1, . . . , x.sub.N-1)H (30)
[0335] where W is the Kronecker product of F.sub.N.sup.-1, which is
the complex conjugate of F.sub.N.sup.-1, and the M.times.M unit
matrix I.sub.M and is represented by the following expression.
W= F.sub.N.sup.-1I.sub.M (31)
[0336] H is a matrix whose diagonal elements are N M.times.M
matrices {tilde over (H)}.sub.i each having the elements of impulse
responses h and IDFT matrices. It is represented by the following
expression.
[ Mathematical expression 22 ] H ~ i = ( h 0 h M - 1 h M - 2 h 1 W
N i _ h 1 h 0 h M - 1 h 2 W N i _ h 2 W N i _ h 1 h 0 h 3 W N i _ h
M - 1 W N i _ h M - 2 W N i _ h M - 3 h 0 ) ( 32 ) [ Mathematical
expression 23 ] H = ( H ~ 0 0 0 0 H ~ 1 0 0 0 H ~ N - 1 ) ( 33 )
##EQU00010##
[0337] where, h.sub.k=0 (L.ltoreq.k.ltoreq.M-1).
[0338] Therefore, if
Y = def ( Y 0 Y 1 Y N - 1 ) Y 0 = def ( Y 00 Y 01 Y 0 ( M - 1 ) ) Y
1 = def ( Y 10 Y 11 Y 1 ( M - 1 ) ) Y N - 1 = def ( Y ( N - 1 ) 0 Y
( N - 1 ) 1 Y ( N - 1 ) ( M - 1 ) ) ##EQU00011##
then, there is the following relation between the transmission data
vectors and the reception signals.
Y.sub.i=x.sub.i{tilde over (H)}.sub.i (0.ltoreq.i.ltoreq.N-1)
(34)
[0339] Let W be a matched filter. Then, when the output Y of the
matched filter and the impulse response h are obtained, solving the
simultaneous equations based on {tilde over (H)}.sub.i gives the
transmission data vectors x.sub.i.
2.3 Features
[0340] The OSDM system has the following features as compared with
the OFDM system.
[0341] (A) The OSDM system, capable of transmitting N transmission
data vectors at the same time, uses one of them for the pilot
signal for measuring impulse responses, providing the communication
path environment with no predictability in real time.
[0342] (B) When estimating the transmission data vectors from the
reception signals, the OSDM system gives the impulse responses of
all communication paths independently. That is, the system allows
the energy of various reflected waves, which arrive at the
receiving side, to be used independently.
[0343] (C) The OFDM system cannot improve the SN ratio in the
frequency area equalization even if the equalization method for
suppressing noises (Minimum Mean Squared Error; MMSE) is used
instead of the equalization method for multiplying the inverse
matrix of an impulse response by the reception signal
(Zero-Forcing; ZF). In contrast, for the reason described above,
the OSDM system can use MMSE to improve the SN ratio.
3. Simulation Result
[0344] This chapter introduces the result of the OSDM system
performance simulation based on the contents of the previous
chapter. Section 1 describes the simulation definitions and Section
2 introduces the simulation result. Finally, Section 3 verifies the
simulation.
3.1 Definitions
[0345] Based on the contents of the previous chapter, the
performance simulation of the OSDM system and the OFDM system was
performed for the baseband signals. The parameters used for the
simulation are as follows.
[0346] M=13
[0347] N=64
[0348] L=8
[0349] The impulse responses on communication paths follow the
independent zero-mean complex Gaussian process, and the signal is
modulated using QPSK and 16 QAM. At this time, the error correction
codes are not used.
[0350] The receiving side adds the Additive White Gaussian Noises
(AWGN) as the thermal noises. In addition, the MMSE equalization is
used as the method for estimating transmission data vectors from
reception signals.
[0351] The experiment method for each system is that the per-bit
power density to noise power density ratio (Eb/No) is changed from
0 [dB] to 25 [dB], 1 [dB] at a time, and Monte Carlo simulation is
carried out 104 times for the Eb/No values.
3.2 Simulation Result
[0352] This section introduces the result of the simulation carried
out based on the definitions in the previous section.
[0353] FIG. 25 and FIG. 26 are graphs showing the per-bit power
density to noise power density ratio (Eb/No) versus the bit error
rate (BER) for the modulation of QPSK and 16 QAM in the OSDM and
OFDM systems. Note that the per-bit power density (Eb) on the
horizontal axis of the graphs includes the energy of not only
direct waves but also all reflected waves. Also, assume that the
receiving side considers the environment of the communication path
as a noise-free, ideal environment.
[0354] Next, FIG. 27 and FIG. 28 show the comparison between the
OSDM system performance when the communication path is estimated as
an ideal communication path and the OSDM system performance when
the communication path is estimated actually from the pilot signal
in the QPSK modulation and in the 16 QAM modulation. In this
experiment, the Zero Correlation Zone (ZCZ) signal having the
length M is used as the pilot signal to estimate the communication
path.
3.3 Investigation
[0355] This section investigates the simulation result introduced
in the previous section.
[0356] First, FIG. 25 and FIG. 26 indicate that the BER
characteristics of the OSDM system are generally better than those
of the OFDM system and that the difference becomes more noticeable
as the per-bit power density to noise power density ratio (Eb/No)
becomes higher. For example, the figures indicate that the OSDM
system achieves the BER of 10.sup.-3 at an Eb/No value that is
about 3 [dB] lower than in the OFDM system. This means that the
OSDM system can provide quality, comparable to that achieved in the
OFDM system, at about one half of the transmission power.
[0357] On the other hand, the BER convergence level in the OFDM
system is not changed much by the modulation, while the convergence
level in the OSDM system tends to get worse only when the Eb/No
value is small as the number of bits per symbol is increased. This
is probably due to the fact that, in an environment where the
signal to noise power ratio (SNR) is low, the effect of the noise
energy at the transmission signal estimation time becomes more
noticeable in the OSDM system than in the OFDM system because the
energy of all reflected waves is independently used in the OSDM
system. However, in a status where the communication path is not
ideally provided, the advantage of the OSDM system is not affected
in an actual communication environment because the Eb/No
deterioration amount of the OSDM system is expected to be smaller
than that of the OFDM system because of the reason that will be
described later.
[0358] FIG. 27 and FIG. 28 indicate that, in the OSDM system, the
Eb/No deterioration amount from the status in which the
communication paths are provided ideally to the status in which the
communication paths are actually estimated is as low as about 3
[dB]. From this result, it is recognized that the pilot signal
correctly provides the receiving side with the communication path
status with no predictability even in a very bad multipath
environment where impulse responses on the communication paths each
follow the independent zero-mean complex Gaussian process. Unlike
the OFDM system in which the preambles, provided for measuring the
communication path environment, are discretely arranged in the
transmission signal, the OSDM system can continuously provide the
communication path environment. Therefore, in a status where the
communication path environment is changed frequently, it is
expected that the communication quality deterioration in the OSDM
system is slighter than that in the OFDM system. In addition,
because the PAPR is almost flat as described in Chapter 1 and
because the ratio of the guard interval to the transmission signal
length is LN+L in the OFDM system but is LMN+L in the OSDM system,
the OSDM system ensures a transmission speed higher than that in
the OFDM system.
4. Theory of Multi-Antenna OSDM System
[0359] This chapter introduces, in four sections, the theory of a
multi-antenna OSDM system, one of applications of the OSDM system,
where data is transmitted and received independently in the same
frequency band using multiple antennas. First, Section 1 introduces
the process of forming transmission signals from data with focus on
the transmission system. Next, Section 2 introduces the process of
creating a communication path environment from reception signals
and estimating data from the created communication path environment
with focus on the receiving system. Finally, Section 3 introduces
the feature of the multi-antenna OSDM system as compared with the
OSDM system.
4.1 Transmission System
[0360] The transmission system is composed of the processes such as
those shown in FIG. 29. The following describes them in
details.
[0361] For the reason defined in Chapter 2, the OSDM system is
characterized in that impulse responses on communication paths can
be obtained with no predictability in real time. That is, the
transmitting side adds special information to the signals so that
impulse responses on multiple communication paths can be correctly
obtained. This configuration allows the OSDM system to be applied
to a multi-antenna communication system where the estimation
accuracy of the communication path environment affects the
communication quality more directly.
[0362] The following describes the transmission and reception over
t antennas using the parameters used in Chapter 2.
[0363] For an antenna i (0.ltoreq.i.ltoreq.t-1), the N-t data
vectors, x.sub.t.sup.i, x.sub.t+1.sup.i, . . . , x.sub.N-1.sup.i,
each of which has the length M, are defined in the same way the
data vectors are defined in (25) as shown in FIG. 30. In addition,
the pilot signal is applied to x.sub.i.sup.i and a zero matrix is
applied to other rows as shown in FIG. 31.
[0364] Next, the Kronecker product of the IDFT matrix and the data
vectors is applied as in Expression (26). Because the Kronecker
product of the pilot signal of transmission data transmitted from
each antenna and one of different rows of the IDFT column is
calculated, the orthogonality of each pilot signal is guaranteed.
That is, it should be noted that, when the transmission signal Si,
generated via Expressions (27) and (28), arrives at the receiving
side via multiple communication paths while being interfered with
other signals, the receiving side can recognize the impulse
responses of the multiple communication paths independently with no
predictability.
4.2 Reception System
[0365] The reception system is configured by the processes such as
those shown in FIG. 32. The following describes the details.
[ Mathematical Expression 24 ] R ~ j = i = 0 t - 1 j = 0 t - 1 k =
0 L - 1 h k i .fwdarw. j S ~ i T k ( 35 ) ##EQU00012##
[0366] The notation shown above is used.
[0367] In the above expression, h.sub.0.sup.i.fwdarw.j,
h.sub.1.sup.i.fwdarw.j, . . . , h.sub.L-1.sup.i.fwdarw.j are
impulse responses on the communication path from the transmission
antenna i to the reception antenna j, and T is the shift matrix
shown in Section 2 of Chapter 2.
[Mathematical Expression 25]
[0368] At this time, the receiving side performs the following
processing.
[0369] First, let R.sup.j be the signal generated by removing the
cyclic prefix from the reception signal {tilde over (R)}.sup.j.
Then,
R=(R.sup.0R.sup.1 . . . R.sup.t-1) (36)
Next, the matched filter W is defined. W is the Kronecker product
of the Nth order DFT matrix FN, N.times.M unit matrix IM, and
t.times.t unit matrix It, and is expressed by the following
expression.
[Mathematical Expression 26]
[0370] W=F.sub.NI.sub.MI.sub.t (37)
[0371] In this case, there is the following relation between the
output RW of the matched filter and transmission data vectors.
Y = def RW = XH Where X = def ( X 0 X 1 X t - 1 ) X 0 = def ( X t 0
0 X t 1 0 X t ( M - 1 ) 0 X ( N - 1 ) ( M - 1 ) 0 ) X 1 = def ( X t
0 1 X t 1 1 X t ( M - 1 ) 1 X ( N - 1 ) ( M - 1 ) 0 ) X t - 1 = def
( X t 0 t - 1 X t 1 t - 1 X t ( M - 1 ) t - 1 X ( N - 1 ) ( M - 1 )
t - 1 ) ( 38 ) ##EQU00013##
[0372] H is a matrix composed of t.sup.2 M.times.M matrices
H.sup.i.fwdarw.j each of which has impulse responses
h.sup.i.fwdarw.j and IDFT matrices as the elements. It is
represented by the following expression.
[ Mathematical Expression 27 ] H ~ k i .fwdarw. j = ( h 0 i
.fwdarw. j h M - 1 i .fwdarw. j h M - 2 i .fwdarw. j h 1 i .fwdarw.
j W N k _ h 1 i .fwdarw. j h 0 i .fwdarw. j h M - 1 i .fwdarw. j h
2 i .fwdarw. j W N k _ h 2 i .fwdarw. j W N k _ h 1 i .fwdarw. j h
0 i .fwdarw. j h 3 i .fwdarw. j W N k _ h M - 1 i .fwdarw. j W N k
_ h M - 2 i .fwdarw. j W N k _ h M - 3 i .fwdarw. j h 0 i .fwdarw.
j ) [ Mathematical Expression 28 ] H ^ i .fwdarw. j = ( H ~ 0 i
.fwdarw. j 0 0 0 H ~ 1 i .fwdarw. j 0 0 0 H ~ N - 1 i .fwdarw. j )
( 39 ) H = ( H ^ 0 .fwdarw. 0 H ^ 0 .fwdarw. 1 H ^ 0 .fwdarw. t
.fwdarw. 1 H ^ 1 .fwdarw. 0 H ^ 1 .fwdarw. 1 H ^ 1 .fwdarw. t
.fwdarw. 1 H ^ t .fwdarw. 1 .fwdarw. 0 H ^ t .fwdarw. 1 .fwdarw. 1
H ^ t .fwdarw. 1 .fwdarw. t .fwdarw. 1 ) ( 40 ) ##EQU00014##
[0373] Let W be a matched filter. Then, when the output Y of the
matched filters and the impulse responses h are obtained, solving
the simultaneous equations based on H gives transmission data
vectors.
4.3 Features
[0374] The multi-antenna OSDM system has the following features as
compared with the multi-antenna OFDM system such as the MIMO-OFDM
system.
[0375] Unlike the OFDM system, the OSDM system gives the
communication path environment with no predictability in real time
as mentioned in Section 3 of Chapter 2. Because the multi-antenna
OSDM system, where t antennas are used for transmission and
reception, reserves a t-row data area as the pilot signal
allocation area, the data area allocated to one antenna is composed
of (N-t) rows. Therefore, as compared with the single-antenna OSDM
system where the pilot signal is allocated to one data area, the
multi-antenna OSDM system is expected to have the information
transmission capacity that is theoretically t(N-t)N-1 times
larger.
[0376] On the other hand, in the MIMO-OFDM system, the transmitting
side and the receiving side share the communication path
environment information and use a method, for example, the
eigenvalue-based beam-forming method, for reserving the
communication path capacity. That is, the comparison of the
transmitting methods themselves indicates that the multi-antenna
OSDM system can reserve a communication path capacity almost
proportional to the number of antennas in a method simpler than
that of the MIMO-OFDM system.
5. Simulation Result
[0377] This chapter introduces the result obtained by simulating
the performance of the multi-antenna OSDM system based on the
contents of the previous chapter. Section 1 describes the
definitions of the simulation and Section 2 introduces the result
of the simulation. Finally, Section 3 describes the
verification.
5.1 Definitions
[0378] Based on the contents of the previous chapter, the
performance simulation of the multi-antenna OSDM system was
performed for the baseband signals.
[0379] The same parameter values as those used in Section 1 of
Chapter 3 are used. The number of transmission/reception antennas
is t, and the impulse responses on the t2 communication paths each
follow the independent zero-mean complex Gaussian process.
[0380] The experiment method is that the per-bit power density to
noise power density ratio (Eb/No) is changed for each system from 0
[dB] to 25 [dB], 1 [dB] at a time, while changing the number of
transmission/reception antennas t (t=1, 2, 4, 8) and, for each
Eb/No value, Monte Carlo simulation is carried out 104 times.
5.2 Simulation Result
[0381] This section introduces the result of the simulation
performed based on the definitions given in the previous section.
FIG. 33 and FIG. 34 are graphs showing the per-bit power density to
noise power density ratio (Eb/No) versus the bit error rate (BER)
in the QPSK modulation and the 16 QAM modulation of the
multi-antenna OSDM system. As in Section 2 of Chapter 3, note that
the per-bit power density used in the graphs includes the energy of
not only direct waves but also all reflected waves in the BER
characteristics in the QPSK modulation in FIG. 23 and in BER
characteristics in the 16 QAM modulation in FIG. 34. Also, assume
that the receiving side considers the environment of the
communication path as a noise-free, ideal environment.
[0382] Next, FIG. 35 and FIG. 36 are graphs showing the number of
transmission/reception antennas versus the throughput in the steady
connection state in the time slot of 1 [.mu.s] when the signal to
noise power ratio (SNR) is 5, 10, and 20 [dB] and when the SNR is
10, 20, and 30 [dB] in the QPSK modulation and the 16 QAM
modulation. Note that the signal power used in the graphs includes
the energy of not only direct waves but also all reflected waves.
The following approximation expression is used to calculate the
throughput.
Throughput.about..alpha..times.(1-BER)/.beta. (41)
[0383] where .alpha. is the number of bits per symbol and .beta. is
the symbol time.
5.3 Verification
[0384] This section verifies the result of the simulation
introduced in the previous section. First, referring to FIG. 33 and
FIG. 34, it is recognized in the multi-antenna OSDM system that the
per-bit power density to noise power density ratio (Eb/No) is
slightly deteriorated even if the number of antennas is increased.
In particular, when Eb/No is high enough, it is important to
remember that the deterioration of Eb/No of the multi-antenna OSDM
system (t=8) is as low as about 3-6 [dB] as compared with that in
the single antenna OSDM system (t=1) introduced in Chapter 2 though
the information transmission capacity is theoretically 7.1 times
larger.
[0385] FIG. 35 and FIG. 36 indicate that, when the signal to noise
power ratio (SNR) is high enough, the multi-antenna OSDM system
(t=8) makes it possible to transmit the information capacity about
seven times larger than that of the single-antenna OSDM system
(t=1). It is recognized that this value is almost equal to the
theoretical value described above.
Multi Virtual Antenna OSDM System
[0386] The following describes an embodiment of multi virtual
antennas.
[0387] The limited and exhaustible wireless frequency resource
becomes a serious issue in wireless communications such as mobile
communications. To address this issue, study has been conducted on
the method for using multiple antennas on both the transmitting
side and the receiving side (MIMO-OFDM) and, by the inventor of the
present invention and his colleagues, on multi-antenna Orthogonal
Signal Division Multiplexing (OSDM). Although multi-antenna OSDM
features wireless frequency usage efficiency much higher than that
of MIMO-OFDM, there is a possibility that the multiple antennas
themselves place a heavy load on a portable communication
device.
[0388] This embodiment, primarily designed for use on a portable
communication device, is based on the "virtual antenna theory"
that, though a single antenna is used both on the transmitting side
and the receiving side, makes possible high wireless frequency
usage efficiency as if multiple antennas were used on both the
transmitting side and the receiving side.
[0389] Because this description is related to the description of
"Technical background of the invention--theory of OSDM" described
above, "multi virtual antenna OSDM system" has the chapter and
section numbers that follow the Chapters and Sections of "Technical
background of the invention--theory of OSDM".
6. Theory of Virtual Antenna on Transmitting Side
[0390] As shown in FIG. 39, it is assumed that there are virtual
transmission antenna #0, virtual transmission antenna #1, . . . ,
virtual transmission antenna #(K-1) and that virtual channel
characteristics #0, virtual channel characteristics #1, . . . ,
virtual channel characteristics #(K-1) are set up such that their
characteristics are different as much as possible.
[0391] Signal #0 is input to virtual transmission antenna #0, . . .
, signal #(K-1) is input to virtual transmission antenna #(K-11),
and the signal generated by adding up the signals passing through
the virtual channels is input to the actual transmission
antenna.
[0392] When the electric wave transmitted from the actual
transmission antenna is received by the actual reception antenna,
signal #0 is affected by the channel characteristics generated by
the convolution between virtual channel characteristics #0 and the
channel characteristics of the actual transmission/reception
antenna because signal #0 passes through the channel between
virtual channel #0 and the actual transmission/reception antenna.
Signal #1, . . . , signal #(K-1) are processed in the same way.
[0393] If the channel characteristics generated by the convolution
between the virtual channel characteristics of the virtual
transmission antennas and the channel characteristics of the actual
transmission/reception antenna are sufficiently different among the
virtual transmission antennas, designing the signals with the
assumption of those virtual transmission antennas will result in
signal #0, . . . , signal #(K-11) being affected by the channel
characteristics as if K transmission antennas and one reception
antenna were used.
6.2 Theory of Virtual Antennas on Receiving Side
[0394] As shown in FIG. 40, K sampling points are set in each time
slot in such a way that the sampling points #0 in the time slots
are at an equal interval, the sampling points #1 in the time slots
are at an equal interval, . . . , and the sampling points #(K-11)
in the time slots are at an equal interval.
[0395] Setting the sampling points as described above will result
in K discrete signals, received from sampling point sequence #0,
sampling point sequence #1, and sampling point sequence #(K-11),
being affected by different channel characteristics as if K
antennas were used and one sampling point was set in each time slot
for each antenna. That is, K virtual reception antennas can be
assumed using one reception antenna.
[0396] As another application of the multi-antenna OSDM system, the
following describes in five sections the theory of the multi
virtual antenna OSDM system in which multiple virtual antennas are
used to transmit and receive data independently in the same
frequency band. First, Section 1 describes the concept of virtual
transmission antennas and Section 2 describes the concept of
virtual reception antennas. Section 3 introduces the process of
forming data from transmission signals with focus on the
transmission system. Next, Section 4 introduces the process of
acquiring the virtual communication path environment from the
received signals and, at the same time, estimating data from the
acquired virtual communication path environment with focus on the
reception system.
6.3 Transmission System
[0397] The transmission system is configured by the processes such
as those shown in FIG. 41. The following describes the details.
[0398] In the description below, t virtual antennas are used for
transmission and reception using the parameters used in Chapter
2.
[0399] For a virtual reception antenna i (0.ltoreq.i.ltoreq.t-1),
N-t data vectors x.sub.t.sup.i, x.sub.t+1.sup.i, . . . ,
x.sub.N-1.sup.i each of which has the length M, such as those shown
in FIG. 30, are defined as in (1). As in FIG. 31, the pilot signal
is applied to x.sub.i.sup.i, and a zero matrix is applied to other
rows.
[0400] Next, the Kronecker product of the IDFT matrix and the data
vectors is applied as in Expression (26). Because the Kronecker
product and the pilot signal of transmission data transmitted from
each virtual transmission antenna and one of different rows of the
IDFT matrix is calculated, the orthogonality of each pilot signal
is guaranteed. That is, it should be noted that, when the
transmission signal Si, generated via Expressions (27) and (28),
arrives at the virtual reception antenna via multiple virtual
communication paths while being interfered with other signals, the
receiving side can recognize the impulse responses of the multiple
virtual communication paths independently with no
predictability.
6.4 Reception System
[0401] The reception system is configured by the process shown in
FIG. 42. The following describes the details.
[ Mathematical Expression 29 ] R ~ j = i = 0 t - 1 j = 0 t - 1 k =
0 L - 1 h k i .fwdarw. j S ~ i T k ( 42 ) ##EQU00015##
[0402] The process can be represented as given above.
h.sub.0.sup.i.fwdarw.j, h.sub.1.sup.i.fwdarw.j, . . . ,
h.sub.L-1.sup.i.fwdarw.j are the impulse responses on the
communication path from the virtual transmission antenna (virtual
transmission antenna) i to the virtual reception antenna j, and T
is the shift matrix shown in Section 2 of Chapter 2.
[0403] The above are the impulse responses on the communication
path from the antenna (virtual transmission antenna) i to the
virtual reception antenna j, and T is the shift matrix shown in
Section 2 of Chapter 2.
[Mathematical Expression 30]
[0404] At this time, the receiving side performs the following
processing. First, let R.sup.j be the signal generated by removing
the cyclic prefix from the received signal {tilde over
(R)}.sup.j.
R=(R.sup.0R.sup.1 . . . R.sup.t-1) (43)
[0405] Next, the matched filter W is defined. Here, W is the
Kronecker product of the Nth order DFT matrix FN, N.times.M unit
matrix IM, and t.times.t unit matrix It, which is shown by the
following expression.
[0406] Next, the matched filter W is defined. Here, W is the
Kronecker product of the Nth order DFT matrix FN, N.times.M unit
matrix IM, and t.times.t unit matrix It, which is shown by the
following expression.
[Mathematical Expression 31]
W=F.sub.NI.sub.MI.sub.t (44)
[0407] In this case, there is the following relation between the
output RW of the matched filter and transmission data vectors.
Y = def RW = XH Where X = def ( X 0 X 1 X t - 1 ) X 0 = def ( X t 0
0 X t 1 0 X t ( M - 1 ) 0 X ( N - 1 ) ( M - 1 ) 0 ) X 1 = def ( X t
0 1 X t 1 1 X t ( M - 1 ) 1 X ( N - 1 ) ( M - 1 ) 0 ) X t - 1 = def
( X t 0 t - 1 X t 1 t - 1 X t ( M - 1 ) t - 1 X ( N - 1 ) ( M - 1 )
t - 1 ) ( 45 ) ##EQU00016##
[0408] H is a matrix composed of t.sup.2 M.times.M matrices
H.sup.i.fwdarw.j each of which has impulse responses
h.sup.i.fwdarw.j and IDFT matrices as the elements. It is
represented by the following expression.
[ Mathematical Expression 32 ] H ~ k i .fwdarw. j = ( h 0 i
.fwdarw. j h M - 1 i .fwdarw. j h M - 2 i .fwdarw. j h 1 i .fwdarw.
j W N k _ h 1 i .fwdarw. j h 0 i .fwdarw. j h M - 1 i .fwdarw. j h
2 i .fwdarw. j W N k _ h 2 i .fwdarw. j W N k _ h 1 i .fwdarw. j h
0 i .fwdarw. j h 3 i .fwdarw. j W N k _ h M - 1 i .fwdarw. j W N k
_ h M - 2 i .fwdarw. j W N k _ h M - 3 i .fwdarw. j h 0 i .fwdarw.
j ) [ Mathematical Expression 33 ] H ^ i .fwdarw. j = ( H ~ 0 i
.fwdarw. j 0 0 0 H ~ 1 i .fwdarw. j 0 0 0 H ~ N - 1 i .fwdarw. j )
( 39 ) H = ( H ^ 0 .fwdarw. 0 H ^ 0 .fwdarw. 1 H ^ 0 .fwdarw. t
.fwdarw. 1 H ^ 1 .fwdarw. 0 H ^ 1 .fwdarw. 1 H ^ 1 .fwdarw. t
.fwdarw. 1 H ^ t .fwdarw. 1 .fwdarw. 0 H ^ t .fwdarw. 1 .fwdarw. 1
H ^ t .fwdarw. 1 .fwdarw. t .fwdarw. 1 ) ( 47 ) ##EQU00017##
[0409] Let W be a matched filter. Then, when the output Y of the
matched filters and the impulse responses h are obtained, solving
the simultaneous equations based on H gives transmission data
vectors.
6.5 Example of Multi Virtual Antenna OSDM when One Transmitting
User is on Transmitting Side and Multiple Receiving Users are on
Receiving Side
[0410] For example, on a downlink in the cellular mobile
communication where there is one transmitting user on the
transmitting side (in some cases, there are multiple actual
antennas for diversity transmission and reception) and multiple
receiving users on the receiving side (typically, there is one
actual antenna for each receiving user but, in some cases, there
may be multiple actual antennas for diversity transmission and
reception), the following must be satisfied.
[0411] Because the receiving side solves the simultaneous
equations, the number of virtual reception antennas of the
receiving user (receiving unit) having one actual antenna must be
larger than or equal to the total number of virtual transmission
antennas from which data is received.
[0412] Because a receiving user, who solves the simultaneous
equations, can receive data transmitted for other receiving users,
the transmission data must be encrypted when transmitted.
6.6 Example of Multi Virtual Antenna OSDM when Multiple
Transmitting Users are on Transmitting Side and One Receiving User
is on Receiving Side
[0413] For example, on an uplink in the cellular mobile
communication where there are multiple transmitting users on the
transmitting side (typically, there is one actual antenna for each
transmitting user but, in some cases, there may be multiple actual
antennas) and one receiving user on the receiving side, the
following must be satisfied.
[0414] In this case, too, because the receiving side solves the
simultaneous equations, the number of virtual reception antennas of
the receiving user (receiving unit) having one actual antenna must
be larger than or equal to the total number of virtual transmission
antennas (number of virtual channels) from which data is
received.
[0415] To prevent interference, the pilot signals of all virtual
transmission antennas must use separate row vectors of those of the
Nth order DFT matrix as the row vectors for the pilot signals.
6.7 Example of Multi Virtual Antenna OSDM when Multiple
Transmitting Users are on Transmitting Side and Multiple Receiving
Users are on Receiving Side
[0416] For example, in the cellular mobile communication where
cell-to-cell interference is caused and there are multiple
transmitting and receiving users on both transmitting side and the
receiving side, separate row vectors of those of the Nth order DFT
matrix must be used as the row vectors for the pilot signals of all
virtual transmission antennas.
[0417] To solve the simultaneous equations on the receiving side,
the total number of virtual reception antennas of one set of
receiving units must be larger than or equal to the total number of
virtual transmission antennas from which data is received.
6.8 Pilot Signal
[0418] It has been described that separate row vectors of those of
the Nth order DFT matrix must be used for the row vectors of the
pilot signals of all virtual transmission antennas.
[0419] However, if a ZCZ sequence set (the sequences are ZACZ
sequences and are ZCCZ sequences with each other) or an approximate
ZCCZ sequence set is used, the same row vector for the pilot signal
may be allocated to multi virtual transmission antennas.
[0420] Which row vector of those of the Nth order DFT matrix is
used for the pilot signal may be set arbitrarily.
7. Verification
[0421] The multi virtual antenna OSDM system was verified under the
following conditions.
[0422] M=13
[0423] N=64
[0424] L=8
[0425] Actual multipaths are Rayleigh fading channels, virtual
multipaths are uniformly-random 16-bit signal channels, and MMSE is
used as the equalization method.
[0426] Eb of EB/No includes not only direct paths but also
reflected paths, and all transmission power transmitted from one
actual antenna is constant.
[0427] The simulation result shown in FIG. 37 and FIG. 38 was
obtained under the conditions given above.
[0428] The present invention can be implemented in the following
modes.
[0429] A transmitting/receiving system wherein, on a transmitting
side, a signal creation unit creates a signal, which is generated
by adding up signals assuming that separate data has passed through
each of a plurality of virtual channels, as an output of the signal
creation unit and, on a receiving side, oversampling is performed,
the sampled data is distributed, and signals are detected assuming
that the distributed data is an output of a plurality of virtual
reception antennas.
[0430] A transmitting device wherein a signal creation unit creates
a signal generated by adding up signals assuming that separate data
has passed through each of a plurality of virtual channels and
[0431] the signal created by the signal creation unit is
transmitted.
[0432] A receiving device wherein oversampling is performed for a
received signal, the sampled data is distributed, and signals are
detected assuming that the distributed data is an output of a
plurality of virtual reception antennas
[0433] A transmitting/receiving system wherein, when separate data
is transmitted from one transmitter to each of a plurality of
receivers, a transmitter side transmits pilot signals in such a way
that the pilot signals can be separated without using channel
characteristics and transmits the separate data by inputting the
separate data into separate virtual transmission channels and
adding up the resulting data and
[0434] each of the plurality of receivers performs oversampling and
distributes the sampled result and, assuming a plurality of virtual
reception antennas, generates plural simultaneous linear equations,
which can estimate transmission data, and estimates transmission
data by solving the plural simultaneous linear equations.
[0435] A receiving device that receives pilot signals and data from
a plurality of transmitters, the pilot signals being transmitted so
that a pilot signal corresponding to each transmitter can be
separated without using channel characteristics, the data being
added to the pilot signals and
[0436] distributes an oversampled result generated by performing
oversampling for received signals and, assuming that the
oversampled result is outputs of a plurality of virtual reception
antennas, separates the transmitters according to channel
characteristics so that plural simultaneous linear equations can
solve diversified channel characteristics between the transmitters
and the receiver, and estimates transmission data.
Technical Meaning of Present Invention
[0437] Shannon showed that the information transmission speed must
not exceed
C=W log.sub.2(S+N/N)
in order to provide the method that makes the error rate as close
to zero as possible, while the present invention shows that "if a
finite error rate is allowed, there is no limit to the information
transmission speed even if the bandwidth is finite".
[0438] The following describes "the method that makes the
information transmission speed infinitely high using a finite
bandwidth".
[0439] Multiple (K) virtual transmission antennas are prepared on
the transmitting side, and signals are generated by the OSDM system
for each virtual transmission antenna using different data. Next,
after inputting the signals to separate virtual channels for each
virtual transmission antenna (convolution of time characteristics),
the signals are added up and transmitted from the actual
transmission antenna.
[0440] On the receiving side, the signal received via the actual
reception antenna is separated by the convolution between the
signal and the time characteristics of separate virtual channels,
and the signals generated by the convolution are output to virtual
reception antennas corresponding to the virtual channels. The
number of virtual reception antennas is K.
[0441] If K virtual transmission antennas and K virtual reception
antennas are used, "the theory of multi antenna OSDM" can be
applied and, regardless of the fact that one actual transmission
antenna and one actual reception antenna are used, the information
transmission speed can be increased about k times that of the
single antenna OSDM (wireless frequency usage efficiency about two
times higher than that of the single antenna OFDM).
[0442] If the amount of calculation and the amount of delay time
need not be considered, K can be increased infinitely. Therefore,
if a finite error rate is allowed, the information transmission
speed can be increased infinitely even in a finite bandwidth.
[0443] Furthermore, OSDM makes it easy to improve the amplitude
distribution of a transmission antenna (prevents the power from
exceeding a predetermined transmission power in a time or frequency
area).
[0444] Although embodiments of the present invention have been
described, it will be understood that the invention is not limited
to the embodiments described above. The present invention may be
changed without departing from the spirit of the invention.
[0445] This international application claims priority from the
prior Japanese Patent Application No. 2007-103078 and
PCT/JP2008/053866, filed Apr. 10, 2007 and Mar. 4, 2008, the entire
contents of Japanese Patent Application No. 2007-103078 and
PCT/JP2008/053866 are incorporated herein by reference.
* * * * *